1. Introduction
Atmospheric rivers (ARs) are a key part of the Californian climate, bringing up to 65% of local seasonal precipitation, principally during the extended winter season (from October to March) (Gershunov et al. 2017). An AR is a long, narrow filament of high integrated vapor transport (IVT), defined by a threshold of 250 kg m−1 s−1 (American Meteorological Society 2019). Landfalling ARs are often accompanied by heavy precipitation, with 92% of the West Coast’s heaviest 3-day rain events fed by ARs (Ralph and Dettinger 2012). Such events can lead to significant hydrometeorological impacts, with economic damage correlated with AR strength and duration (Corringham et al. 2019).
Atmospheric rivers most often develop over the data-sparse regions of midlatitude oceans, where the primary sources of observational data are from satellites and buoys. The scientific research on ARs has been increasing rapidly over the past decade (Ralph et al. 2017a), and with it, increasing interest and demand for observations. Dropsonde measurements of ARs in the Northeast Pacific used in this study began with the CalWater project in 2014 and 2015 (Ralph et al. 2016), and continued with Atmospheric River Reconnaissance (AR Recon) in 2016, 2018, 2019 and 2020 (Ralph et al. 2020). These observational campaigns involve the deployment of targeted airborne and buoy observations over the Northeast Pacific from January to March. CalWater focused on observing the structure of ARs (Cordeira et al. 2017), while AR Recon’s primary goal is to improve forecasts of the landfall and impacts of ARs on the U.S. West Coast at lead times of 1–5 days. In addition, in winter 2015–16 the NASA-led Olympic Mountain Experiment (OLYMPEX) took place on the Olympic Peninsula of Washington State, with the goal of evaluating satellite-derived precipitation measurements, sampling atmospheric rivers among other storm systems (Houze et al. 2017).
The key observations used in this study are from dropsondes, which record pressure, temperature, wind, and relative humidity at very high vertical resolution throughout the atmosphere as they descend from the aircraft to the sea surface, which is important as observing system research has shown the need to sample the whole atmospheric column, not just the surface meteorology (Ralph et al. 2014). Data from dropsondes are assimilated by global modeling centers in real time with the aim of improving the initial conditions. After experience from earlier seasons, AR Recon was designated as an “operational” requirement, as directed by the National Winter Season Operations Plan (OFCM 2019) and adopted the methods developed in 2016–19 field campaigns. The 2020 field season of AR Recon more than doubled the number of research flights compared with what had been done in 2016, 2018, and 2019, combined. These occurred on 17 dates centered on 0000 UTC, ±6 h [each of these dates is referred to as an “intensive observation period” (IOP)] and involved 1–3 aircraft and 20–80 dropsondes per IOP.
Flight paths and dropsonde locations are derived by quantitatively pinpointing locations of greatest sensitivity in the forecast to AR landfall IVT and precipitation with 1–5 days lead time. Adjoint sensitivity using the Coupled Ocean–Atmosphere Mesoscale Prediction System (COAMPS) (Hodur 1997) indicates both low-level winds and precipitation are most sensitive to mid- to lower-tropospheric perturbations in the initial state in and near the ARs, with largest sensitivity to moisture, followed by temperature and winds (Reynolds et al. 2019). Alongside adjoint methods, ensemble sensitivity (e.g., Torn and Hakim 2008) using the Global Ensemble Forecast System, the Canadian Meteorological Center, and the European Centre for Medium-Range Weather Forecasts models helps to inform targeted observation locations. A key goal of AR reconnaissance campaigns is to gather observations that reduce forecast error and uncertainty in real time, therefore providing time for appropriate impact mitigation to take place.
Not only are observational data assimilated into operational forecast models, but they are also used in research studies to further understand the dynamics and processes that are the main drivers of key AR characteristics. In addition, they are vital for model verification, including on reanalyses (e.g., Guan et al. 2018) and operational models (e.g., Lavers et al. 2018, 2019). However, the observations only represent a single column and so representativeness errors need to be considered when comparing these point observations to model data (Lavers et al. 2018).
Studies such as ours, which focus on the research applications of observational data (e.g., dropsonde measurements) can help to inform and improve the processes in which these observations are gathered, by highlighting physical mechanisms important to downstream precipitation and/or areas with high forecast uncertainty.
Over the U.S. West Coast, 82% of ARs are associated with an extratropical cyclone, centered over the strongest pressure gradient region with a low over the northwest and a high over the southeast (Zhang et al. 2019). ARs are often described as forming part of the broader warm conveyor belt of extratropical cyclones (e.g., Eckhardt et al. 2004), although an AR is an Earth-relative airflow and a warm conveyor belt is a cyclone-relative airflow (Dacre et al. 2019). A feature of these systems is the pre-cold-frontal low-level jet, which is due to the strong horizontal temperature gradient at the surface cold front causing the geostrophic wind to decrease with height, resulting in a jet maximum at lower levels (Browning and Pardoe 1973). It has also been reported that ARs have a moist-neutral vertical distribution of equivalent potential temperature (Ralph et al. 2005; Neiman et al. 2013b) throughout their approximate 3 km depth (Ralph et al. 2004, 2017b; Guan and Waliser 2017). This near-neutral moist static stability, along with large moisture transport by a low-level jet can lead to intense orographic precipitation along coastal and windward mountain ridges (Ralph et al. 2004, 2005, 2006; Neiman et al. 2008a). Orographic precipitation involves a complex interaction between mountain geometry (Hughes et al. 2014), thermodynamics and cloud microphysics (Morales et al. 2018), with three-dimensional distributions of water vapor flux and orientation relative to topography important for precipitation extremes associated with ARs (Hughes et al. 2014; Hecht and Cordeira 2017). The layer of upslope flow that optimally modulates orographic rainfall is about 1 km above mean sea level for California’s coastal range, corresponding to the mean altitude of landfalling low-level jets (Neiman et al. 2002).
Cases in which ARs occur without an extratropical cyclone are often in the lower latitudes (<35°N) as there is sufficient moisture (Zhang et al. 2019), and primarily occur during the summer months (Knippertz et al. 2018). This analysis focuses on ARs in the winter and excludes tropical systems; therefore, the data examined are mostly associated with extratropical cyclones with low pressure to the northwest and high pressure to southeast. The existence of an extratropical cyclone, however, is not a requirement for the events in this study, nor is the relationship between ARs, extratropical cyclones, and fronts explored here.
Many previous studies have relied on much smaller observational datasets (e.g., Ralph et al. 2005, 2017b), on satellite-derived products (e.g., Ralph et al. 2004; Neiman et al. 2008a,b) or on reanalyses or model data (e.g., Leung and Qian 2009; Wick et al. 2013; Rutz et al. 2014) to examine the characteristics of atmospheric rivers. This study involves the analysis of 858 vertical profiles of temperature, wind and moisture from 33 observational events. As the aims of the CalWater and AR Recon campaigns are to further knowledge of ARs, and dropsondes are primarily released in and around them, the interpretation of this study is limited to regions of high IVT. Here, we use dropsondes that form transects across these regions of elevated IVT (i.e., IVT > 250 kg m−1 s−1), which we consider to be ARs. We define AR sectors based on location relative to maximum IVT for a given transect and explore the mean and variance of composite vertical profiles. We also present in-depth analysis of the low-level jet, moist neutrality and water vapor transport vertical distribution in relation to these IVT-based sectors. In section 2, a description of the observational data is presented along with a method for identifying sectors of the AR. Results are discussed in section 3 and conclusions are presented in section 4.
2. Data and methods
a. Observational data
The dropsonde data have been collected as part of CalWater (Ralph et al. 2016) and AR Recon (Ralph et al. 2020), with the aim of sampling ARs and areas of greatest forecast sensitivity to U.S. West Coast precipitation (Reynolds et al. 2019). There are a total of 1251 dropsonde observations from 33 IOPs over 6 winter seasons (January–March) (2014, 2015, 2016, 2018, 2019, 2020) (Table 1, Fig. 1a), of which 858 are analyzed. Some of these dropsondes are excluded from this study when assigning atmospheric river sector (see section 2e). Several other IOPs have been excluded from this analysis as there are fewer than 5 dropsondes within the AR (250 kg m−1 s−1 threshold) or the peak IVT of the system is in the tropics (below 23.5°N). Previous work has defined ARs as objects with length > 2000 km, and a length/width ratio of > 2 (e.g., Guan and Waliser 2015; Guan et al. 2018). Here we have used IVT derived from ERA5 reanalysis data (described below) to visually confirm that all remaining IOPs examined in this study sample ARs, with long (>2000 km) filaments of high IVT (>250 kg m−1 s−1), although we have not explicitly imposed a length/width ratio.
Details of the reconnaissance campaigns, intensive observing period (IOPs), flights, and dropsondes used in this analysis. Time (UTC) reflects the time that the dropsondes are centered around. A total of 33 IOPs and 1251 dropsondes.
(a) Release location of all 1251 dropsondes (black dots) used in this study. Observations collected from IOPs during 2014, 2015, 2016, 2018, 2019, and 2020. Mean IVT calculated from ERA5 across central time of these 33 IOPs shown in colored contours. (b) Distribution of drift distance (km) from 6 km altitude to 25 m for 858 dropsondes that meet the AR sector identification criteria detailed in section 2e.
Citation: Monthly Weather Review 149, 3; 10.1175/MWR-D-20-0177.1
(a) Release location of all 1251 dropsondes (black dots) used in this study. Observations collected from IOPs during 2014, 2015, 2016, 2018, 2019, and 2020. Mean IVT calculated from ERA5 across central time of these 33 IOPs shown in colored contours. (b) Distribution of drift distance (km) from 6 km altitude to 25 m for 858 dropsondes that meet the AR sector identification criteria detailed in section 2e.
Citation: Monthly Weather Review 149, 3; 10.1175/MWR-D-20-0177.1
(a) Release location of all 1251 dropsondes (black dots) used in this study. Observations collected from IOPs during 2014, 2015, 2016, 2018, 2019, and 2020. Mean IVT calculated from ERA5 across central time of these 33 IOPs shown in colored contours. (b) Distribution of drift distance (km) from 6 km altitude to 25 m for 858 dropsondes that meet the AR sector identification criteria detailed in section 2e.
Citation: Monthly Weather Review 149, 3; 10.1175/MWR-D-20-0177.1
The dropsondes are generally released in a 6-h window, and since 2016 this has been from 2030 to 0230 UTC day +1, centered around 0000 UTC for the purposes of assimilation into forecast models. Prior to 2016 there was no requirement to collect data in a particular assimilation window as there were no real-time forecasts using this dropsonde data.
The original dropsonde used was the Vaisala RD94, and since 2019 this has since been replaced by the Vaisala RD41 in both NOAA (R. Henning 2020, personal communication) and U.S. Air Force AR operations (Lt. Col. Ryan Rickert 2020, personal communication). Both the RD94 and RD41 dropsondes record pressure, temperature, and humidity twice, and meridional and zonal components of wind four times a second as they descend through the atmospheric column (Vaisala 2017, 2018). With a descent speed of approximately 11 m s−1 at sea level and 21 m s−1 at 12 km altitude, the vertical resolution is approximately 5–10 m and the last measured pressure level is approximately 5–6 m above the surface (Vaisala 2018). The resolution of pressure, temperature, and relative humidity measurements as well as the repeatability of temperature have improved with the newer RD41 model (Table 2).
Resolution and repeatability of RD41 and RD94 dropsondes. Repeatability refers to a standard deviation of differences between two successive repeated calibrations, k = 2 confidence level. [Source: Vaisala (2017, 2018).
The geopotential altitude, which integrates pressure, temperature and humidity using the hypsometric equation from the surface upward, is used in this study. Depending on the fall speed of the dropsonde, the last partial frame measured before hitting the surface is likely between 5 and 10 m above the surface and the Atmospheric Sounding Processing Environment (Aspen) (www.eol.ucar.edu/content/aspen) program takes this into account and makes a best estimate of the geopotential height of the last recorded data frame, with an uncertainty of approximately ±3 m (H. Vömel 2020, personal communication). Both the upper and lower limits of the vertical profile vary between dropsondes, and only those that record temperature, humidity and pressure from 25 m above the surface and extend to 6000 m in altitude are analyzed further in this study. This range spans the depth of a typical AR, with 75% water vapor transport below 3 km (American Meteorological Society 2019) and approximately 5% above 6 km (Ralph et al. 2017a).
Profiles where vertical gaps in data exceed 150 m are removed. Remaining data are linearly interpolated onto every meter so that density of observations in the vertical does not skew the composite plots. The mean horizontal distance traveled from 6000 m to 25 m altitude is 11.5 km, with a range of 0 to 21.5 km (Fig. 1b). This distance is comparable to an average global model gridbox size and can be considered as sampling a single column of air. This drift distance is calculated by taking the difference between the location at 6000 m and that at 25 m and so does not account for variations within this layer. For example, the dropsonde that shows a drift value of 0 km over this depth drifts within the column 3.3 km away from the initial position at 6000 m, but drifts back to its initial latitude and longitude location at 25 m altitude. Figure 1a shows the release locations of all of the dropsondes used in this analysis with the mean IVT from ERA5 (Hersbach et al. 2020) at the central times of each IOP examined.
b. IVT calculation
c. Planetary boundary layer height calculation
Using the bulk Richardson number and threshold of 0.25 is a widely used method to calculate the PBL height, although the determination of the critical value is rather insensitive for mixed layers with a capping inversion (Seidel et al. 2012).
d. Brunt–Väisälä frequency calculation
To determine whether the dry (
e. Atmospheric river sector identification
The dropsondes are classified into five different sectors based on IVT across individual transects (Table 3). For each flight track, a transect is defined where the dropsondes are deployed in a line that crosses the main AR axis and samples the highest IVT (visually determined using IVT calculated from ERA5 data), e.g., see Fig. 2. Flights that do not meet this criterion are not analyzed in this study, reducing the sample size from 1251 to 858 dropsondes. IVT was calculated for all dropsondes, and those with less than 250 kg m−1 s−1 (threshold for determining the presence of an AR) were assigned as non-AR. In each transect, the dropsonde with the largest IVT was identified and the percentage of this IVT value was calculated for all other dropsondes within the transect. The AR core is defined as IVT within a critical percentage of the maximum IVT, and sensitivity to three different core thresholds was tested using 75%, 80%, and 85%. The aim was to define AR sectors that were of approximate equal width, and this was achieved with the 80% threshold. The results of this study were not found to be sensitive to the choice of core percentage (not shown), and so only those using this 80% threshold are described further, with the core drops having ≥80% max IVT and >250 kg m−1 s−1. The remaining dropsondes within the AR (IVT > 250 kg m−1 s−1) and outside of the core were assigned AR sectors. The dropsondes equatorward of the AR core outer limit were labeled as non-AR warm side or AR warm sector, and those poleward were labeled as non-AR cold side or AR cold sector (Table 3). Not all transects sampled all five sectors described.
AR sector category definitions.
Locations of dropsondes deployed during IOP5 2020, centered around 0000 UTC 5 Feb. Color of dots reflects sector of dropsonde (see legend): non-AR cold side (NCS), AR cold sector (CS), AR core (C), AR warm sector (WS), and non-AR warm side (NWS). See Table 3 for category details. ERA5 IVT shown in colored contours and mean sea level pressure shown in gray contour lines for 0000 UTC 5 Feb.
Citation: Monthly Weather Review 149, 3; 10.1175/MWR-D-20-0177.1
Locations of dropsondes deployed during IOP5 2020, centered around 0000 UTC 5 Feb. Color of dots reflects sector of dropsonde (see legend): non-AR cold side (NCS), AR cold sector (CS), AR core (C), AR warm sector (WS), and non-AR warm side (NWS). See Table 3 for category details. ERA5 IVT shown in colored contours and mean sea level pressure shown in gray contour lines for 0000 UTC 5 Feb.
Citation: Monthly Weather Review 149, 3; 10.1175/MWR-D-20-0177.1
Locations of dropsondes deployed during IOP5 2020, centered around 0000 UTC 5 Feb. Color of dots reflects sector of dropsonde (see legend): non-AR cold side (NCS), AR cold sector (CS), AR core (C), AR warm sector (WS), and non-AR warm side (NWS). See Table 3 for category details. ERA5 IVT shown in colored contours and mean sea level pressure shown in gray contour lines for 0000 UTC 5 Feb.
Citation: Monthly Weather Review 149, 3; 10.1175/MWR-D-20-0177.1
Summary of AR sector identification method:
Identify that flight track crosses the AR axis with at least five dropsondes.
Calculate IVT for each dropsonde.
Assign dropsondes with <250 kg m−1 s−1 IVT to non-AR sides (see Table 3).
For remaining dropsondes, identify that with highest IVT in the transect.
Calculate percentage of maximum IVT for dropsondes in transect.
Use this percentage to allocate dropsondes to AR sectors (see Table 3).
Figure 2 shows the ERA5 IVT and dropsonde locations for IOP5 2020, color-coded by sector according to the above definition. Broadly, the core (black dots) follows the most intense central IVT band of the AR. Equatorward and within the colored contours (250 kg m−1 s−1 threshold) are the magenta warm sector dropsondes. Poleward of the core drops and within the 250 kg m−1 s−1 threshold are the cold sector dropsondes (cyan dots) and further poleward still are the non-AR cold side drops (blue dots). If any dropsonde in the core of a given transect begins below 6000 m, terminates above 25 m, or has gaps in measurements larger than 150 m, the entire transect is removed from analysis as an accurate measurement of maximum IVT cannot be obtained. This manual quality control resulted in a total of 78 transects available to use in this study. Differences between the ERA5 and dropsonde IVT are apparent in some cases, for example, where dropsondes assigned to non-AR sectors fall within the 250 kg m−1 s−1 colored contour. Such differences between the observational and reanalysis IVT are not the focus of this study and will be addressed in following work.
A recent study by Guan et al. (2020) defined four AR sectors using an AR object boundary and the location of maximum IVT across several transects. The ARs were detected by an algorithm based on a combination of geometry and intensity thresholds (Guan and Waliser 2015; Guan et al. 2018), which identifies transects and defines AR objects. These sectors (postfrontal sector, frontal, prefrontal and pre-AR) are based on the understanding of the typical association of a midlatitude AR with a frontal zone and the movement of this synoptic system. The method defined in this study is a complementary approach that does not require the definition of an AR object nor the existence of a frontal zone. While there are additional criteria that could be used to identify sectors of an AR, such as thresholds of temperature or winds, our approach here is AR relative, relying on just the IVT field. It is made to be simple and allows us to explore AR characteristics and variability by compositing several hundreds of dropsondes into these different sectors. We are also taking advantage of a large dataset of in situ observations rather than relying on model output.
3. Results
A total of 78 transects from 33 IOPs have been examined, with dropsondes divided into sectors based on an 80% core threshold (see section 2e). The number of dropsondes, alongside IVT and width distributions are shown in Fig. 3. AR strength is also shown on Fig. 3b, determined by IVT and categorized into weak, moderate, strong, extreme, and exceptional as in Ralph et al. (2019). The majority of core IVT values are moderate to strong (500–1000 kg m−1 s−1), with outliers extending to the lower end of the weak category (~300 kg m−1 s−1) and to the border of the extreme and exceptional categories (~1250 kg m−1 s−1) (Fig. 3b). The AR sector widths are determined using transects that fully sample at least one sector, giving sample sizes of 47, 66, and 30 and means of 268, 285, and 290 km, for the cold sector, core and warm sector, respectively (Fig. 3a). Results show that there are large standard deviations of more than 40% of the width itself (Table 4), and given the variety in AR shape and orientation, this large variation in sector width is perhaps unsurprising. The box-and-whisker plots (Fig. 3a) reveal a largely Gaussian distribution of AR width in the core and warm sector and a positively skewed distribution for the AR cold sector. Widths of the non-AR sides cannot be obtained as there are no boundaries to these sectors; however, the mean distance of dropsondes in these sides to the dropsonde with maximum IVT in the transect are 559 ± 252 km for the non-AR cold side and 511 ± 169 km for the non-AR warm side. The mean total AR width is obtained by summing the widths of each AR sector, and the standard deviation is calculated by taking the square root of the variances of each AR sector width. The resultant mean total AR width is 831 ± 278 km, which is comparable to 890 ± 270 km reported in Ralph et al. (2017b). The research flights are not always exactly perpendicular to the AR. To compensate, all width and distance values are based on the component of the flight path perpendicular to the AR mean IVT direction.
(a) Bar chart showing the number of dropsondes in each sector: non-AR cold side (NCS), cold sector (CS), core (C), warm sector (WS), and non-AR warm side (NWS). Box-and-whisker plot of AR sector width (CS, C, WS) on secondary/right y axis. (b) Box-and-whisker plot of dropsonde IVT calculated for each sector. AR strength is labeled on the secondary/right y axis (weak, moderate, strong, extreme, exceptional) according to Ralph et al. (2019). For all box-and-whisker plots, brown horizontal lines are the median, gray dashed horizontal lines are the mean, the box is the interquartile range (Q1–Q3), and whiskers extend from 5th to 95th percentiles, with outliers shown in dots.
Citation: Monthly Weather Review 149, 3; 10.1175/MWR-D-20-0177.1
(a) Bar chart showing the number of dropsondes in each sector: non-AR cold side (NCS), cold sector (CS), core (C), warm sector (WS), and non-AR warm side (NWS). Box-and-whisker plot of AR sector width (CS, C, WS) on secondary/right y axis. (b) Box-and-whisker plot of dropsonde IVT calculated for each sector. AR strength is labeled on the secondary/right y axis (weak, moderate, strong, extreme, exceptional) according to Ralph et al. (2019). For all box-and-whisker plots, brown horizontal lines are the median, gray dashed horizontal lines are the mean, the box is the interquartile range (Q1–Q3), and whiskers extend from 5th to 95th percentiles, with outliers shown in dots.
Citation: Monthly Weather Review 149, 3; 10.1175/MWR-D-20-0177.1
(a) Bar chart showing the number of dropsondes in each sector: non-AR cold side (NCS), cold sector (CS), core (C), warm sector (WS), and non-AR warm side (NWS). Box-and-whisker plot of AR sector width (CS, C, WS) on secondary/right y axis. (b) Box-and-whisker plot of dropsonde IVT calculated for each sector. AR strength is labeled on the secondary/right y axis (weak, moderate, strong, extreme, exceptional) according to Ralph et al. (2019). For all box-and-whisker plots, brown horizontal lines are the median, gray dashed horizontal lines are the mean, the box is the interquartile range (Q1–Q3), and whiskers extend from 5th to 95th percentiles, with outliers shown in dots.
Citation: Monthly Weather Review 149, 3; 10.1175/MWR-D-20-0177.1
Summary table of characteristics of sectors within an atmospheric river. Number of dropsondes in each sample shown in parentheses within the table for those that do not use the full sample labeled in the header. Mean and standard deviation are reported for all variables, apart from low-level jet height (mode) and low-level jet percentage occurrence. All low-level jet results are from a sample determined using a threshold depth of 350 m above maximum wind speed to find wind speed 2 m s−1 weaker. Width and total IVT (TIVT) are calculated using dropsonde data when at least one sector is sampled in the transect. Planetary boundary layer (PBL), lifting condensation level (LCL), and convective available potential energy (CAPE).
The remainder of this analysis uses 154 dropsondes in the non-AR cold side, 197 in the cold sector, 247 in the core, 207 in the warm sector, and 53 in the non-AR warm side (Fig. 3a). The total number of dropsondes is 858, reduced from the 1251 detailed in Table 1 following the criteria laid out in section 2e, including transect definitions and quality control of full depth data.
Total IVT (TIVT) is an “AR transport” metric, analogous to the streamflow in a terrestrial river. It is the horizontal integral of IVT across the AR transect perpendicular to the direction of mean vapor transport and the Ralph et al. (2017b) study reported a mean value of 4.7 × 108 ± 2 × 108 kg s−1. In our study, using transects in which at least one AR sector is fully sampled (i.e., both edges defined), the mean TIVT calculated is 4.40 × 108 ± 1.65 × 108, with 46% in the core, 29% in the cold sector and 25% in the warm sector. Using this definition of AR sectors described in section 2e gives approximate symmetry in terms of TIVT either side of the AR core. Additionally, the AR sector definition presented in this study relies solely on IVT, the principal measure of AR strength, and is a simple method that can be easily applied across models and observations without the need for additional data or algorithms.
a. Composite vertical profiles
Composite vertical profiles using 154 non-AR cold side (NCS), 197 cold sector (CS), 247 core (C), 207 warm sector (WS), and 53 non-AR warm side (NWS) dropsondes. (a) Wind speed (m s−1), (b) U (m s−1), and (c) V (m s−1). The 95% confidence interval is shown at 500 m increments.
Citation: Monthly Weather Review 149, 3; 10.1175/MWR-D-20-0177.1
Composite vertical profiles using 154 non-AR cold side (NCS), 197 cold sector (CS), 247 core (C), 207 warm sector (WS), and 53 non-AR warm side (NWS) dropsondes. (a) Wind speed (m s−1), (b) U (m s−1), and (c) V (m s−1). The 95% confidence interval is shown at 500 m increments.
Citation: Monthly Weather Review 149, 3; 10.1175/MWR-D-20-0177.1
Composite vertical profiles using 154 non-AR cold side (NCS), 197 cold sector (CS), 247 core (C), 207 warm sector (WS), and 53 non-AR warm side (NWS) dropsondes. (a) Wind speed (m s−1), (b) U (m s−1), and (c) V (m s−1). The 95% confidence interval is shown at 500 m increments.
Citation: Monthly Weather Review 149, 3; 10.1175/MWR-D-20-0177.1
Within the AR, stronger wind speeds are observed compared to the non-AR sectors, with the core exhibiting the highest wind speeds up to approximately 4000 m (Fig. 4a). The maximum mean wind speed below 1000 m is approximately 24 m s−1 in the core, 19 m s−1 in the warm sector, and 16 m s−1 in the cold sector. The core value compares well to the mean wind speed at 1000 m of 23.4 m s−1 found in a 17 dropsonde composite in Ralph et al. (2005), which examined profiles that exhibited a low-level jet. In these composite profiles a low-level jet is apparent in the warm sector and to some extent in the core, and this feature will be explored in more detail in section 3b. The confidence intervals around wind speed in Ralph et al. (2005) were approximately 16 m s−1 compared to less than 1 m s−1 shown here in most sectors. This is due to the much larger sample sizes of 154, 197, 247, 207, and 53 (see Fig. 3) compared to the 17 in the aforementioned study [see Eq. (7)].
At low levels (below 1000 m), all sectors apart from the non-AR cold side have a stronger meridional than zonal component (Figs. 4b,c). Composite profiles for the core, warm sector and non-AR warm side show relatively constant meridional wind above 1000 m but with increasing zonal wind throughout the 6000 m layer. This vertical shear characteristic of warm advection is apparent in the mean wind direction (Fig. 5a) as wind veers with height. Warm advection is further investigated, and Figs. 5b and 5c show the percentage occurrence and strength of warm advection for all sectors. The presence of warm advection was determined over 50 m depths, using the 5 m mean at the upper and lower levels. For example, in the lowest level the mean wind direction over 25–30 m is subtracted from the mean wind direction over 75–80 m, and this is repeated for 119 levels up to 6000 m. At low levels (below 1000 m), the core and warm sector dropsondes show the greatest occurrence of warm advection, from ~75% at the surface to ~50% at 1000 m, with all sectors exhibiting a large decrease in occurrence of warm advection over this depth (Fig. 5b). The process of warm air advection is forcing for ascent, conducive to the formation of clouds and precipitation, which are most prevalent in the core and warm sector. The magnitude of the warm advection shown by the cumulative mean wind veer shows largely similar values in the lowest 1000 m for all sectors, but with some divergence above this level, particularly of the non-AR sectors (Fig. 5c). The magnitude of wind veering from the surface to 1000 m in the core is approximately 50° (Fig. 5c), and this can be compared to ~30° in Fig. 5a when all profiles (including those that do not have warm advection) are taken into account. The mean wind direction of the non-AR cold side is ~260° at the surface compared to ~200° for all other sectors. This reflects the fact that the air mass forming the non-AR cold side originates from a more westerly direction compared to the other sectors, which have a more southerly origin.
(a) Mean wind direction (degrees) using 154 non-AR cold side (NCS), 197 cold sector (CS), 247 core (C), 207 warm sector (WS), and 53 non-AR warm side (NWS) dropsondes. (b) Percentage occurrence of warm advection for all sectors calculated over 50 m depths. (c) Cumulative mean wind veer (degrees) of profiles that exhibit warm advection.
Citation: Monthly Weather Review 149, 3; 10.1175/MWR-D-20-0177.1
(a) Mean wind direction (degrees) using 154 non-AR cold side (NCS), 197 cold sector (CS), 247 core (C), 207 warm sector (WS), and 53 non-AR warm side (NWS) dropsondes. (b) Percentage occurrence of warm advection for all sectors calculated over 50 m depths. (c) Cumulative mean wind veer (degrees) of profiles that exhibit warm advection.
Citation: Monthly Weather Review 149, 3; 10.1175/MWR-D-20-0177.1
(a) Mean wind direction (degrees) using 154 non-AR cold side (NCS), 197 cold sector (CS), 247 core (C), 207 warm sector (WS), and 53 non-AR warm side (NWS) dropsondes. (b) Percentage occurrence of warm advection for all sectors calculated over 50 m depths. (c) Cumulative mean wind veer (degrees) of profiles that exhibit warm advection.
Citation: Monthly Weather Review 149, 3; 10.1175/MWR-D-20-0177.1
The dropsondes analyzed in this study are mainly over the open ocean (Fig. 1). However, 14 are within 100 km of land and 3 are within 50 km (26, 35, 33 km) of land. Of these, 12 (out of 14) and 2 (out of 3) are off the U.S. West Coast, with the remaining dropsondes in the vicinity of Hawaii. The steep orography and associated barrier jets along the U.S. West Coast have been shown to influence ARs, in particular redistributing precipitation at landfall (Smith et al. 2010; Neiman et al. 2013a). Yu and Smull (2000) examined the effect of terrain-blocked flow upstream of steep coastal orography of Oregon and northern California and found that effects were confined to 20 km offshore. With all dropsondes analyzed more than 20 km offshore, we can assume that in this study, any influence of steep orography on wind speed and direction is negligible.
The vertical composites of potential temperature (θ) show a largely stable environment across all sectors with a relatively constant vertical gradient of 4.8 to 5.3 K km−1 for the AR sectors (Fig. 6a), which is comparable to the dry static stability of 4.5 K km−1 found in Ralph et al. (2005). However, throughout this 6000 m layer there is significant moisture, especially in the core, and particularly at low levels (Figs. 6c,d). Since stability depends critically on the saturation of the air, equivalent potential temperature is a more appropriate metric to examine.
Composite vertical profiles using 154 non-AR cold side (NCS), 197 cold sector (CS), 247 core (C), 207 warm sector (WS), and 53 non-AR warm side (NWS) dropsondes. (a) Potential temperature (K), (b) equivalent potential temperature (K), (c) relative humidity (%), and (d) water vapor mixing ratio (g kg−1). The 95% confidence interval is shown as horizontal bars at 500 m increments.
Citation: Monthly Weather Review 149, 3; 10.1175/MWR-D-20-0177.1
Composite vertical profiles using 154 non-AR cold side (NCS), 197 cold sector (CS), 247 core (C), 207 warm sector (WS), and 53 non-AR warm side (NWS) dropsondes. (a) Potential temperature (K), (b) equivalent potential temperature (K), (c) relative humidity (%), and (d) water vapor mixing ratio (g kg−1). The 95% confidence interval is shown as horizontal bars at 500 m increments.
Citation: Monthly Weather Review 149, 3; 10.1175/MWR-D-20-0177.1
Composite vertical profiles using 154 non-AR cold side (NCS), 197 cold sector (CS), 247 core (C), 207 warm sector (WS), and 53 non-AR warm side (NWS) dropsondes. (a) Potential temperature (K), (b) equivalent potential temperature (K), (c) relative humidity (%), and (d) water vapor mixing ratio (g kg−1). The 95% confidence interval is shown as horizontal bars at 500 m increments.
Citation: Monthly Weather Review 149, 3; 10.1175/MWR-D-20-0177.1
The composites of equivalent potential temperature (θe) show more distinct profiles (Fig. 6b) than potential temperature between the sectors. The cool and dry non-AR cold side exhibits the lowest θe throughout the 6000 m shown and has a similar shaped vertical profile to the non-AR warm side, both with a negative gradient in the lowest 1500 m and a positive gradient above this level (Fig. 6b). This change in behavior is highlighted in the composites of relative humidity, where the non-AR cold side and non-AR warm side have a sharp negative gradient in the lowest 1500 m, with less variability in relative humidity in the upper 4000 m or so (Fig. 6c). The warm sector θe profile has a minimum at approximately 2000 m, but with a less defined inflection point as the relative humidity decreases more smoothly with height. The core and cold sector do not exhibit this negative θe gradient in the lowest 1500–2000 m. Instead, the cold sector remains relatively stable at approximately 307.5 K and the core exhibits a constant gradient between 1000 and 2000 m, but with a positive gradient in the lowest 1000 m, which is indicative of a moist stable air mass.
The presence of moist neutral conditions in the core of the AR below 2.5 km has been documented in previous studies (Ralph et al. 2005; Neiman et al. 2008a; Conrick and Mass 2019) and this is identified by Sawyer (1956) as one of the conditions favorable for orographic enhancement of rainfall, as it allows for the complete lifting of oncoming air by the obstacle. In a composite of 17 soundings, Ralph et al. (2005) showed largely constant θe gradient with height at ~311 K with a confidence interval of ~11 K. The occurrence of moist neutral profiles in the AR sectors is further explored in section 3c.
Consistent with the location of the AR core relative to the maximum IVT, the water vapor mixing ratio and relative humidity are generally greater in the core compared to all other sectors (Figs. 6c,d), remaining above 80% humidity through the lowest 3000 m. The core composite water vapor mixing ratio profile compares well to that reported in Neiman et al. (2002), who showed moist conditions (nearly 10 g kg−1) below 300 m and a steady decrease of moisture with height aloft. Interestingly, the relative humidity profile in the cold sector compares favorably to the core, while the warm sector is similar to the non-AR sides. The IVT distributions (Fig. 3b) and moisture content above ~1500 m (Fig. 6d) are comparable between the cold sector and warm sector. Therefore, this difference in relative humidity is likely explained by the difference in temperature; the warm sector is on average ~5 K warmer throughout the 6000 m layer (Fig. 6a).
b. Low-level jet
Although the composite profiles of wind in each sector appear distinct (Fig. 4a), they are not separated by one standard deviation, which is shown in Fig. 7a for the AR sectors (cold sector, core, and warm sector). The spread is due to both differences in the magnitude and the profile shape among the >150 dropsondes in each AR sector, with low-level jets apparent in some cases. We identify the presence of a low-level jet by locating a wind maximum below 1500 m residing beneath a wind speed at least 2 m s−1 weaker (e.g., Neiman et al. 2002; Ralph et al. 2005). Several layer depths were used to find the weaker winds aloft, from 150 to 600 m at 50 m increments above the level of the maximum wind. We find an increase in the percentage occurrence of the low-level jet with increasing layer depth (Fig. 7b). The core, warm sector, and non-AR warm side have higher frequencies of low-level jets than the cold sector and non-AR cold side. We expect to find the low-level jet in the core and warm sector as these are most likely to sample the pre–cold front environment. The comparable frequency of low-level jets in the non-AR warm side is the result of processes outside of the AR environment. Using limits that define a shallow low-level jet, i.e., weaker winds aloft within 150 m of the wind maximum, low-level jets are apparent in ~15% of the core profiles and ~20% of the warm sector profiles. Raising this depth to 350 m increases the percentage occurrence to ~50% and ~55%, and over 600 m, to ~65% and ~72%, in the core and warm sector, respectively.
(a) Composite vertical profile of wind speed and one standard deviation in shading for the AR sectors: cold sector (CS), core (C), warm sector (WS), with 197, 247, 207 profiles, respectively. (b) Percentage occurrence of low-level jet between 100 and 1500 m for all sectors. The y axis is the depth above the wind maximum used to locate a wind 2 m s−1 weaker (from 150 to 600 m, at 50 m increments).
Citation: Monthly Weather Review 149, 3; 10.1175/MWR-D-20-0177.1
(a) Composite vertical profile of wind speed and one standard deviation in shading for the AR sectors: cold sector (CS), core (C), warm sector (WS), with 197, 247, 207 profiles, respectively. (b) Percentage occurrence of low-level jet between 100 and 1500 m for all sectors. The y axis is the depth above the wind maximum used to locate a wind 2 m s−1 weaker (from 150 to 600 m, at 50 m increments).
Citation: Monthly Weather Review 149, 3; 10.1175/MWR-D-20-0177.1
(a) Composite vertical profile of wind speed and one standard deviation in shading for the AR sectors: cold sector (CS), core (C), warm sector (WS), with 197, 247, 207 profiles, respectively. (b) Percentage occurrence of low-level jet between 100 and 1500 m for all sectors. The y axis is the depth above the wind maximum used to locate a wind 2 m s−1 weaker (from 150 to 600 m, at 50 m increments).
Citation: Monthly Weather Review 149, 3; 10.1175/MWR-D-20-0177.1
The distribution of the depth of the low-level jet from the level of the wind maximum to the level aloft where winds are 2 m s−1 weaker is shown for the AR sectors (Fig. 8), and is remarkably similar across all AR sectors. This figure shows the results using low-level jets obtained by searching for the wind minimum aloft over 5 depths (150, 250, 350, 450, 550 m) above the wind maximum. With increasing layer depth, the upper range (i.e., 95th-percentile whiskers and outliers) is extended and the lower level remains almost constant as these lower depths have already been accounted for. In the largest layer depth (550 m), the means (gray dashed lines) are larger than the medians (brown lines) for all AR sectors, indicating a skewed distribution with tails at the high end. Such a non-Gaussian distribution is less apparent using a 350 m depth limit, with means and medians more closely aligned in all sectors, with values of between approximately 175 and 200 m. Limiting the search for 2 m s−1 weaker winds aloft to 150 m above the maximum negatively skews the distribution, with tails at the low end. The interquartile range for all distributions remains below 340 m and the mean depth ranges from 100 m with the lowest local range of 150 m, to ~250 m when searching over 550 m for the weaker winds (Fig. 8). Therefore, the following height and wind speed analysis uses the low-level jets determined by searching 250, 300 and 350 m above the wind maximum.
Depth of the low-level jet in the AR sectors: cold sector (CS), core (C), warm sector (WS), with sample size annotated above each bar. Brown horizontal lines are medians, gray dashed horizontal lines are the means, the box is the interquartile range (Q1–Q3), and whiskers extend from 5th to 95th percentiles with outliers shown as dots.
Citation: Monthly Weather Review 149, 3; 10.1175/MWR-D-20-0177.1
Depth of the low-level jet in the AR sectors: cold sector (CS), core (C), warm sector (WS), with sample size annotated above each bar. Brown horizontal lines are medians, gray dashed horizontal lines are the means, the box is the interquartile range (Q1–Q3), and whiskers extend from 5th to 95th percentiles with outliers shown as dots.
Citation: Monthly Weather Review 149, 3; 10.1175/MWR-D-20-0177.1
Depth of the low-level jet in the AR sectors: cold sector (CS), core (C), warm sector (WS), with sample size annotated above each bar. Brown horizontal lines are medians, gray dashed horizontal lines are the means, the box is the interquartile range (Q1–Q3), and whiskers extend from 5th to 95th percentiles with outliers shown as dots.
Citation: Monthly Weather Review 149, 3; 10.1175/MWR-D-20-0177.1
The distributions of maximum jet wind speed and height where this occurs are shown for low-level jet profiles determined using 250, 300, and 350 m from level of wind maximum to 2 m s−1 weaker winds aloft, with transparency increasing with depth (Figs. 9a,b). Using the 350 m depth, the height of the low-level jet in the core is most frequently found at approximately 522 m, compared to ~520 m in the cold sector and ~416 m in the warm sector (Fig. 9a) with means of 764 ± 370, 770 ± 389, and 633 ± 346 m, respectively (Table 4). The cold sector exhibits a larger spread in the height of the jet maximum, with greater occurrence above 800 m than the other AR sectors, which suggests that jets in this sector, as defined here, are controlled by different processes due to their likely location behind the cold front. There have been several previous studies that have identified the presence of a low-level jet in observations of ARs and ahead of the cold front of extratropical cyclones, which find a low-level jet at approximately 1 km (Neiman et al. 2002; Ralph et al. 2005; Demirdjian et al. 2020), 900–850 hPa (Browning and Pardoe 1973), and 800–850 hPa (Lackmann 2002). The regions in this study that are most closely related to these previous findings are the core and warm sector, where the mean heights of the low-level jets are 764 and 633 m, which are much lower than previously reported. The presence of a cold front is not determined in this current study, as identification of sector is based solely on IVT. Therefore, if a cold front is present, it is possible that the dropsondes assigned to the core are representative of regions both ahead of and behind the cold front. The mean width of the core is 283 ± 115 km (Fig. 3a, Table 4), which is a relatively large area compared to those low-level jet studies that focused on a smaller defined area in relation to a front. The warm sector has a mean width of 290 ± 199 km (Fig. 3a, Table 4) and in some cases dropsondes may sample just ahead of a cold front if the core is narrow, but will also sample regions distant from the cold front. Any signal of a low-level jet ahead of a cold front is likely to be diluted by the nature of how these sectors have been determined, using IVT and not a front related metric. This study has analyzed many more vertical profiles than these previous studies and the large standard deviations highlight the variability in the low-level jet height.
(a) Distribution of the height of the maximum wind speed (m) in the low-level jets diagnosed using three different depth ranges to find the 2 m s−1 weaker winds aloft (250, 300, 350 m, with increasing transparency). AR sectors are shown: cold sector (CS), core (C), and warm sector (WS). (b) As in (a), but distribution of maximum wind speed in all sectors, that is, including non-AR cold side (NCS) and non-AR warm side (NWS). Dots indicate means using the low-level jets diagnosed with the 350 m depth range above wind speed maximum to find the 2 m s−1 weaker winds aloft.
Citation: Monthly Weather Review 149, 3; 10.1175/MWR-D-20-0177.1
(a) Distribution of the height of the maximum wind speed (m) in the low-level jets diagnosed using three different depth ranges to find the 2 m s−1 weaker winds aloft (250, 300, 350 m, with increasing transparency). AR sectors are shown: cold sector (CS), core (C), and warm sector (WS). (b) As in (a), but distribution of maximum wind speed in all sectors, that is, including non-AR cold side (NCS) and non-AR warm side (NWS). Dots indicate means using the low-level jets diagnosed with the 350 m depth range above wind speed maximum to find the 2 m s−1 weaker winds aloft.
Citation: Monthly Weather Review 149, 3; 10.1175/MWR-D-20-0177.1
(a) Distribution of the height of the maximum wind speed (m) in the low-level jets diagnosed using three different depth ranges to find the 2 m s−1 weaker winds aloft (250, 300, 350 m, with increasing transparency). AR sectors are shown: cold sector (CS), core (C), and warm sector (WS). (b) As in (a), but distribution of maximum wind speed in all sectors, that is, including non-AR cold side (NCS) and non-AR warm side (NWS). Dots indicate means using the low-level jets diagnosed with the 350 m depth range above wind speed maximum to find the 2 m s−1 weaker winds aloft.
Citation: Monthly Weather Review 149, 3; 10.1175/MWR-D-20-0177.1
The low-level jet intensities found in Browning and Pardoe (1973), Lackmann (2002), Neiman et al. (2002), Ralph et al. (2005, 2017b), and Demirdjian et al. (2020) are largely between 20 and 30 m s−1, in the same range reported here in the AR sectors (Table 4), with some wind speeds exceeding 35 m s−1, which is seen in the core in this study. The low-level jets outside of the AR, in the non-AR cold and warm sides, are weaker than within the AR (Fig. 9b), with mean wind speeds of 15.7 ± 6.1 m s−1 and 13.5 ± 4.9 m s−1, respectively. If a wind speed threshold was imposed on this low-level jet identification method, there would be reduced occurrences outside of the AR sectors. The low-level jet in the core exhibits the highest wind speed, with a mean (using the 350 m depth range) of 28.3 ± 6.4 m s−1 (Table 4) with some values extending to 45 m s−1. The cold sector and warm sector have similar distributions of jet maximum wind speed with means of 22.7 ± 5.6 m s−1 and 21.9 ± 5.2 m s−1. Variability has been found around these jet speeds, in relation to distance from the coast (Neiman et al. 2002), latitude of system (Ralph et al. 2017b) and also with large-scale climate variability such as El Niño–Southern Oscillation (ENSO) (Ralph et al. 2005). In the sample we analyze here, there is one weak La Niña event (2018), three weak El Niño events (2015, 2019, 2020), and one very strong El Niño event (2016) (ONI dataset accessible from https://origin.cpc.ncep.noaa.gov/products/analysis_monitoring/ensostuff/ONI_v5.php). Ralph et al. (2005) found that there were stronger low-level jet winds associated with an AR in the strong El Niño of winter 1998, than in an AR during a weak La Niña in 2001. The effect of ENSO conditions could contribute to the variability of the low-level jet observed here, although the limited number of ENSO cases does not allow for a robust analysis of these aspects. The relative importance of the wind field in producing an AR increases with latitude, with increasing low-level jet wind speeds observed in the midlatitudes (>33°N) compared to the subtropics (<33°N) (Ralph et al. 2017b). Tropical ARs, where the peak IVT of the system is below 23.5°N were excluded from this study, and all remaining dropsondes were treated as a whole to generate the composite picture of AR sectors. The effect of latitude on the low-level jet could introduce some variability in these results, with the latitude of dropsondes spanning close to 40° (Fig. 1). As mentioned in section 3a, the vast majority of dropsondes are more than 50 km away from land and so any topographical influence on the jet is likely to be negligible in this study.
When ARs are associated with extratropical storms, there is often the presence of a low-level jet due to the strong temperature gradient observed at the cold front. However, extratropical storms are not always present in the vicinity of ARs, and future work will subset dropsondes based on whether or not the AR is accompanied by an extratropical cyclone, to assess how their presence may influence the results presented here.
c. Moist static stability
Moist static stability of approximately zero (i.e., neutral) is a characteristic that has been found in ARs and pre-cold-frontal low-level jets in the lowest 2.5 km (Ralph et al. 2005). Moist neutral conditions have been identified as one of the conditions favorable for orographic enhancement of rainfall, allowing for the complete lifting of oncoming air by the obstacle (Sawyer 1956). In the composite vertical profiles of equivalent potential temperature (Fig. 6b) this characteristic (i.e., constant θe with height) is apparent in the cold sector between the surface and 2000 m, and in the core between 1000 and 4000 m. Figure 10a shows mean θe and one standard deviation for the AR sectors up to 4000 m, with large spread observed in all sectors.
(a) Composite vertical profiles of equivalent potential temperature and one standard deviation in shading for the AR sectors: cold sector (CS), core (C), warm sector (WS), with 197, 247, 207 profiles, respectively. (b) Mean vertical displacement to reach saturation for AR sectors. The 95% confidence interval is shown at 500 m increments. The 400 m displacement to reach saturation is marked as a dotted gray line. (c) Mean squared moist Brunt–Väisälä frequency (
Citation: Monthly Weather Review 149, 3; 10.1175/MWR-D-20-0177.1
(a) Composite vertical profiles of equivalent potential temperature and one standard deviation in shading for the AR sectors: cold sector (CS), core (C), warm sector (WS), with 197, 247, 207 profiles, respectively. (b) Mean vertical displacement to reach saturation for AR sectors. The 95% confidence interval is shown at 500 m increments. The 400 m displacement to reach saturation is marked as a dotted gray line. (c) Mean squared moist Brunt–Väisälä frequency (
Citation: Monthly Weather Review 149, 3; 10.1175/MWR-D-20-0177.1
(a) Composite vertical profiles of equivalent potential temperature and one standard deviation in shading for the AR sectors: cold sector (CS), core (C), warm sector (WS), with 197, 247, 207 profiles, respectively. (b) Mean vertical displacement to reach saturation for AR sectors. The 95% confidence interval is shown at 500 m increments. The 400 m displacement to reach saturation is marked as a dotted gray line. (c) Mean squared moist Brunt–Väisälä frequency (
Citation: Monthly Weather Review 149, 3; 10.1175/MWR-D-20-0177.1
To determine if the assessment of moist static stability is appropriate, the vertical displacement required to reach saturation has been calculated at 10 m increments for each profile, and the mean and 95% confidence interval for each AR sector is shown (Fig. 10b). With the height of the coastal mountains at 500–1500 m, we conservatively assume that when less than 400 m lift is required to reach saturation, the profile is saturated and assessment of moist stability is valid. The mean vertical displacement required for the core to reach saturation is below 400 m up to approximately 2800 m, which is line with the result in Ralph et al. (2005). The cold and warm sectors require significantly more than 400 m lift to reach saturation above ~500 m and therefore moist stability in these regions will only be assessed below ~500 m.
The mean squared moist Brunt–Väisälä frequency (hereafter BVF) (
On average the troposphere is stable and the average BVF (Nd) is ~0.01 s−1 (Vallis 2017), i.e.,
Occurrence of moist neutral stability or reduced stability in the core, using squared moist Brunt–Väisälä frequency (
Citation: Monthly Weather Review 149, 3; 10.1175/MWR-D-20-0177.1
Occurrence of moist neutral stability or reduced stability in the core, using squared moist Brunt–Väisälä frequency (
Citation: Monthly Weather Review 149, 3; 10.1175/MWR-D-20-0177.1
Occurrence of moist neutral stability or reduced stability in the core, using squared moist Brunt–Väisälä frequency (
Citation: Monthly Weather Review 149, 3; 10.1175/MWR-D-20-0177.1
In the lowest layer (25–525 m), where the cold sector and warm sector can be assessed using this moist metric, there are zero profiles satisfying the 0.1 × 10−4 s−2 threshold (not shown), as in the core. For the remaining thresholds using the 90% criterion, values range from 2%–17% in the cold sector and 1%–11% in the warm sector (not shown), compared to 6%–19% in the core.
Figure 11c shows the mean occurrence of core squared moist BVF values within the various thresholds in each 500 m layer. On average, such values meet the moist neutral thresholds of 0.1 and 0.2 × 10−4 s−2 ~20%–30% and ~55%–60% of the time. Using the less conservative thresholds, these mean frequencies increase to ~75%. Interestingly, examining the mean frequency does not show an increase with height as in Fig. 11a.
Alongside examining relatively shallow layers of 500 m depth, we present results for the core where the top level varies between 2000 and 3000 m and the lower level increases from 25 to 1000 m at 25 m increments (Figs. 11b,e). Approximately zero core profiles meet the squared moist BVF thresholds of 0.3 × 10−4 s−2 and below, 90% of the time (not shown). The highest percentage occurrence of profiles meeting the 0.4 × 10−4 s−2 and 0.5 × 10−4 s−2 thresholds are in the smallest layer depth, of 1000–2000 m, with ~20% and ~24%, respectively (Fig. 11b). Throughout the depth of the AR (surface to 3000 m), approximately 8% and 12% of profiles in the core have squared moist BVF values between 0–0.4 × 10−4 s−2 and 0–0.5 × 10−4 s−2 90% of the time (Fig. 11b), indicating reduced stability compared to normal.
Figure 11e uses deviation of θe from the mean to assess occurrence of moist neutrality, using thresholds of 0.75 and 1 K, as values were zero for smaller thresholds (not shown). Approximately 7.5% of the core profiles meet the 0.75 K threshold 90% of the time between 1000 and 2000 m, and raising this upper level to 3000 m reduces the percentage occurrence to ~3%. Increasing the θe threshold to 1 K results in almost 28% in the 1000 to 2000 m layer, and 17% in the 1000–3000 m layer.
Examination of the squared moist BVF (
Using 17 dropsondes in the low-level jet, Ralph et al. (2005) found relatively constant mean θe from surface to 2500 m, although with a large confidence interval of approximately ±5 K. With our larger sample size, we also find a relatively constant mean θe in the core, but only above 1000 m, with the existence of moist stability (i.e., θe increasing with height) in the lowest 1000 m, which has not been previously observed (Figs. 6b and 10a). Moist stability of the core is assessed using the squared moist BVF, in which stable layers are defined if
The lifting condensation level (LCL) is the height at which an air parcel would saturate if lifted adiabatically, and distributions for all sectors are shown in Fig. 12a. As with the IVT distributions (Fig. 3), the LCL distributions for all sectors are significantly different from each other at the 99% confidence level, except for the warm sector and cold sector, and the non-AR warm side and non-AR cold side. The LCL 95th percentiles for the core and warm sector are below 500 m, which is the lower end of the scale height of California’s coastal mountains (500–1500 m) that would provide the orographic lift. Therefore, it can be assumed that 95% of the air approaching the coast in the core and warm sector will be orographically lifted to saturation. The cold sector has a similar median to the warm sector (~220 m), but with a positive skewed distribution with a tail toward higher values. These LCL results show that the core and the warm sector are preconditioned to saturate at the coastal ranges, resulting in orographic precipitation. The process of intense orographic precipitation due to the vertical displacement of a moist neutral atmosphere with little resistance to lifting is explored in model simulations by Miglietta and Rotunno (2005), who find that there is a possibility of nonlinear switching between statically neutral and stable conditions considering that stability depends on air saturation, which depends on parcel displacement, which depends on stability. It should be noted that we observe these stability characteristics in profiles mostly located over the open ocean (see Fig. 1a) and therefore, are likely to change as the AR approaches the California coastal range.
(a) Box-and-whisker plot of lifting condensation level (LCL) (m). Brown horizontal lines are medians, gray dashed horizontal lines are means, the box is the interquartile range (Q1–Q3), and whiskers extend from 5th to 95th percentiles with outliers shown as dots. (b) Cumulative count of CAPE values for all dropsondes in each sector: 154 non-AR cold side (NCS), 197 cold sector (CS), 247 core (C), 207 warm sector (WS), and 53 non-AR warm side (NWS) dropsondes.
Citation: Monthly Weather Review 149, 3; 10.1175/MWR-D-20-0177.1
(a) Box-and-whisker plot of lifting condensation level (LCL) (m). Brown horizontal lines are medians, gray dashed horizontal lines are means, the box is the interquartile range (Q1–Q3), and whiskers extend from 5th to 95th percentiles with outliers shown as dots. (b) Cumulative count of CAPE values for all dropsondes in each sector: 154 non-AR cold side (NCS), 197 cold sector (CS), 247 core (C), 207 warm sector (WS), and 53 non-AR warm side (NWS) dropsondes.
Citation: Monthly Weather Review 149, 3; 10.1175/MWR-D-20-0177.1
(a) Box-and-whisker plot of lifting condensation level (LCL) (m). Brown horizontal lines are medians, gray dashed horizontal lines are means, the box is the interquartile range (Q1–Q3), and whiskers extend from 5th to 95th percentiles with outliers shown as dots. (b) Cumulative count of CAPE values for all dropsondes in each sector: 154 non-AR cold side (NCS), 197 cold sector (CS), 247 core (C), 207 warm sector (WS), and 53 non-AR warm side (NWS) dropsondes.
Citation: Monthly Weather Review 149, 3; 10.1175/MWR-D-20-0177.1
The distribution of convective available potential energy (CAPE) for dropsondes in all sectors is shown in Fig. 12b, and in all sectors, the majority of dropsondes have values of less than 10 J kg−1. The core and cold sector contain dropsondes with the highest values of CAPE (~340 J kg−1), although the frequency of such values is very small. The mean CAPE values of the AR sectors are 28.4, 13.8 and 9.8 J kg−1 for the cold sector, core and warm sector, respectively (Table 4). These relatively low values are comparable to a study by Oakley et al. (2017) who found that the median CAPE during a historic AR that produced extreme precipitation over California was 20–40 J kg−1, and reflect the fact that ARs are not highly convective systems.
d. Water vapor transport vertical distribution
The water vapor transport is a product of specific humidity and horizontal wind speed. Across all sectors, there is a maximum in mean water vapor transport at approximately 500–600 m (Fig. 13a) with values in the cold sector, core and warm sector of approximately 112, 214, and 170 g kg−1 m s−1, respectively. The height of the largest flux is in the vicinity of the low-level wind jet maximum (section 3b), which was also found in Ralph et al. (2005).
(a) Composite vertical profiles of water vapor flux using 154 non-AR cold side, 197 cold sector, 247 core, 207 warm sector, and 53 non-AR warm side dropsondes. The 95% confidence interval is shown at 500 m increments. (b) Fraction of IVT with height for AR sectors (CS, C, WS). Mean is shown as a solid line and one standard deviation is shown as shading. The 0.5 and 0.75 fractions are shown as dashed vertical lines with the corresponding height value marked as triangles and stars on the y axis. (c) Cumulative IVT with mean in solid line and one standard deviation in shading for all sectors.
Citation: Monthly Weather Review 149, 3; 10.1175/MWR-D-20-0177.1
(a) Composite vertical profiles of water vapor flux using 154 non-AR cold side, 197 cold sector, 247 core, 207 warm sector, and 53 non-AR warm side dropsondes. The 95% confidence interval is shown at 500 m increments. (b) Fraction of IVT with height for AR sectors (CS, C, WS). Mean is shown as a solid line and one standard deviation is shown as shading. The 0.5 and 0.75 fractions are shown as dashed vertical lines with the corresponding height value marked as triangles and stars on the y axis. (c) Cumulative IVT with mean in solid line and one standard deviation in shading for all sectors.
Citation: Monthly Weather Review 149, 3; 10.1175/MWR-D-20-0177.1
(a) Composite vertical profiles of water vapor flux using 154 non-AR cold side, 197 cold sector, 247 core, 207 warm sector, and 53 non-AR warm side dropsondes. The 95% confidence interval is shown at 500 m increments. (b) Fraction of IVT with height for AR sectors (CS, C, WS). Mean is shown as a solid line and one standard deviation is shown as shading. The 0.5 and 0.75 fractions are shown as dashed vertical lines with the corresponding height value marked as triangles and stars on the y axis. (c) Cumulative IVT with mean in solid line and one standard deviation in shading for all sectors.
Citation: Monthly Weather Review 149, 3; 10.1175/MWR-D-20-0177.1
Figure 13b shows the fraction of IVT with height and reveals that 0.5 and 0.75 are reached in the core below 1801 and 3115 m, occurring in between the heights for the cold sector and warm sector. These mean heights and standard deviation for AR sectors are shown in Table 4. Across all AR sectors, half of the total IVT occurs in the lowest ~1300–2000 m and 75% below approximately 2700–3400 m. Previous literature suggests that 75% of the water vapor flux occurs below 2.25 km (Ralph et al. 2005) or 3 km (Ralph et al. 2017b), with the latter study showing slightly lower height on the warm side of the AR and higher on the cold side, consistent with these findings (see Table 4), which also show that the low-level jet is lower in the warm sector (section 3b). It is perhaps surprising that the height of the low-level jet in this study is lower than in previous studies and the height at which 75% IVT occurs is higher than previously observed. However, in this study we quantify the variability around the jet height and the depth at which this IVT amount is detected, which has not previously been documented using such a large sample size. It should be noted that the total depth over which IVT is calculated in this study is up to 6000 m (~476 hPa, section 2b), unlike 300 hPa used in other studies, although approximately only 5% of IVT is above 6000 m, as reported in Fig. 6 of Ralph et al. (2017a). The cumulative IVT is shown in Fig. 13c and reveals that the non-AR sides (NCS and NWS) have very similar distributions in the 25–6000 m layer. The warm sector and cold sector have almost identical mean total IVT of approximately 450 kg m−1 s−1 (also seen in Fig. 3), but with different vertical distributions. The mean core IVT is approximately 750 kg m−1 s−1, with one standard deviation from the core, warm sector, and cold sector means overlapping at approximately 600 kg m−1 s−1.
The distribution of planetary boundary layer (PBL) height is presented in Fig. 14a and shows the median and mean for all sectors falling between 644 and 835 m, and interquartile ranges spanning 450 to 1200 m. The PBL height distribution in the core is significantly different from all other sectors at the 95% (or greater) confidence level. The warm sector and non-AR warm side also significantly differ from each other at the 95% confidence level. All other sector combinations, however, are not significantly different from one another. Most dropsondes in this analysis are deployed in the 6 h centered over 0000 UTC (1600 PST). The PBL height exhibits diurnal variation in temperature and winds, however, this is muted over the oceans, which have a high heat content compared to land.
(a) Box-and-whisker plots of planetary boundary layer (PBL) height, calculated using the bulk Richardson number. For all box-and-whisker plots, brown horizontal lines are medians, gray dashed horizontal lines are means, the box is the interquartile range (Q1–Q3), and whiskers extend from 5th to 95th percentiles with outliers shown as dots. (b) Scatterplot of PBL height against lifting condensation level (LCL) with regression lines for each sectors. The R values for each sector are shown in the legend. (c) Distribution of the fraction of IVT at the PBL height for AR sectors. Colored triangles mark the modes of the distributions. (d) Distribution of the difference between the height below which 75% of the IVT is found (H75) and the PBL height (m) for AR sectors. Dots mark the means of the distributions.
Citation: Monthly Weather Review 149, 3; 10.1175/MWR-D-20-0177.1
(a) Box-and-whisker plots of planetary boundary layer (PBL) height, calculated using the bulk Richardson number. For all box-and-whisker plots, brown horizontal lines are medians, gray dashed horizontal lines are means, the box is the interquartile range (Q1–Q3), and whiskers extend from 5th to 95th percentiles with outliers shown as dots. (b) Scatterplot of PBL height against lifting condensation level (LCL) with regression lines for each sectors. The R values for each sector are shown in the legend. (c) Distribution of the fraction of IVT at the PBL height for AR sectors. Colored triangles mark the modes of the distributions. (d) Distribution of the difference between the height below which 75% of the IVT is found (H75) and the PBL height (m) for AR sectors. Dots mark the means of the distributions.
Citation: Monthly Weather Review 149, 3; 10.1175/MWR-D-20-0177.1
(a) Box-and-whisker plots of planetary boundary layer (PBL) height, calculated using the bulk Richardson number. For all box-and-whisker plots, brown horizontal lines are medians, gray dashed horizontal lines are means, the box is the interquartile range (Q1–Q3), and whiskers extend from 5th to 95th percentiles with outliers shown as dots. (b) Scatterplot of PBL height against lifting condensation level (LCL) with regression lines for each sectors. The R values for each sector are shown in the legend. (c) Distribution of the fraction of IVT at the PBL height for AR sectors. Colored triangles mark the modes of the distributions. (d) Distribution of the difference between the height below which 75% of the IVT is found (H75) and the PBL height (m) for AR sectors. Dots mark the means of the distributions.
Citation: Monthly Weather Review 149, 3; 10.1175/MWR-D-20-0177.1
Figure 14b shows positive correlations between the PBLH and LCL in all sectors, with the strongest correlations in the cold sector and non-AR cold side (R = 0.74 and R = 0.71) and the weakest in the core (R = 0.36), with significance at p < 0.01 in all sectors. In the absence of forced ascent, the LCL and PBL height could be linked through the presence of low-level clouds; a low LCL indicates large amounts of moisture within the layer and a higher likelihood of low-level clouds, which can in turn limit the growth of the boundary layer, resulting in a lower PBL height, and vice versa. For a given PBL height, a different LCL is observed across the sectors, the lowest in the core, followed by the AR sectors, then the non-AR sectors. Where the LCL and hence PBL height are lower, there is a greater likelihood of reaching saturation when lifted by the coastal mountains, leading to orographic precipitation. Where the LCL and PBL height are higher than the coastal mountains, the chance of orographic precipitation is decreased.
For every dropsonde, the cumulative percentage of IVT at the PBL height was calculated and its distribution for each AR sector is shown in Fig. 14c. The mode of each distribution reveals that the boundary layer most frequently contains ~0.08, ~0.16, and ~0.19 of the total IVT in the cold sector, core and warm sector, respectively. The skewed distribution highlights the large spread of the fraction of IVT contained within the PBL, spanning a range of over 60%. Figure 14d combines results found in Figs. 13b and 14a and quantifies the difference between the PBL height and that below which 75% of the IVT is contained (H75). Results show that 75% of the IVT is reached approximately 2500–3000 m above the PBL for all AR sectors.
Although the Bulk Richardson number has deficiencies [e.g., it is insensitive for mixed layers with a capping inversion (Seidel et al. 2012)], it is sufficient for the results shown here. There are other methods for determining the PBL height, but with only three variables available from the dropsondes and no direct measure of vertical velocity, this method is most appropriate for our use. Future work will calculate vertical velocity from the dropsondes and conduct a more detailed study on variations of PBL height between AR sectors.
4. Conclusions
The analysis presented here examines 858 vertical profiles across 78 atmospheric river (AR) transects collected during 33 IOPs. The large sample size allowed for a robust analysis on the characteristics found in 5 different AR sectors: non-AR cold side (NCS), cold sector (CS), core (C), warm sector (WS), and non-AR warm side (NWS). The AR core is defined as having 80% of the maximum integrated vapor transport (IVT) across each transect. Dropsondes with IVT less than 250 kg m−1 s−1 (threshold for determining the presence of an AR) were assigned as non-AR, and those within the AR (IVT > 250 kg m−1 s−1) and outside of the core were assigned AR sectors, with those equatorward of the AR core labeled as non-AR warm side or AR warm sector, and those poleward labeled as non-AR cold side or AR cold sector. The sensitivity of the results to the core threshold, measured as percentage of maximum IVT in the transect was tested using 75% and 85% and there was negligible difference in the results presented here. Choosing a core threshold of 80% provided comparable sample sizes and similar widths of all AR sectors.
Although the cold sector, core and warm sector were all of similar width, we showed that approximately 50% of the total IVT (TIVT) is in the core, with ~25% in the cold and warm sectors. The total IVT in this study of 4.42 × 108 kg s−1, is comparable to 4.7 × 108 kg s−1 reported in Ralph et al. (2017b).
The composite vertical profiles of wind speed show larger values in the AR compared to outside of the AR up to 4000 m, and warm advection in all sectors. The core and warm sector have the greatest occurrence of warm advection, with wind veering with height in ~75% of dropsonde profiles at the surface. The low-level (below 1000 m) maximum mean wind speed is approximately 24 m s−1 in the core, 19 m s−1 in the warm sector and 16 m s−1 in the cold sector and although the 95% confidence interval is small (<1 m s−1) due to a large sample size, there is still a large standard deviation due to difference in wind magnitude and profile shape.
Using hundreds of vertical profiles across the AR allows for robust assessment of the low-level jet and quantification of variability in height, intensity, and depth. Understanding of the low-level jet is important for rainfall, as Neiman et al. (2002) found that orographic precipitation efficiency was 50% greater when a low-level jet was present than when one was not. Low-level jet analysis revealed the strongest jet in the core, with a mean of 28.3 ± 6.4 m s−1 which is within the range found in previous studies (Lackmann 2002; Ralph et al. 2017b). The most frequent occurrence of low-level jets was found in the core and warm sector, and although they were present outside of the AR, they were significantly weaker. The mean height of the low-level jet was 770 m in the cold sector, 764 m in the core and 633 m in the warm sector, which are all lower than the approximate 1 km height reported in previous observational papers (Neiman et al. 2002; Ralph et al. 2005; Demirdjian et al. 2020). This could be due to a larger area across the AR being sampled rather than pinpointing the exact location of the cold front, as well as the larger sample size analyzed. A lower low-level jet height has implications for orographic precipitation efficiency, as Neiman et al. (2002) found that upslope flow 1 km above mean sea level optimally modulates orographic rainfall at California’s coastal range. The depth of the low-level jet from the wind speed maximum to winds 2 m s−1 weaker aloft showed similar distributions across all AR sectors. Mean depths of the low-level jet range from ~100 to ~250 m and the interquartile range for all distributions remains below 340 m.
Throughout all sectors there is significant moisture present, especially in the AR and at low levels, so although dry static stability indicates very stable conditions, a more appropriate metric is moist static stability. In the lowest 2000 m, the warm sector exhibits a negative θe vertical gradient and the cold sector exhibits almost constant θe with height. The θe profile in the core shows a different behavior, with a relatively constant θe between 1000 and 2000 m, and a positive vertical gradient from the surface to 1000 m, which is indicative of a moist stable air mass. However, squared moist BVF analysis showed no core profiles exceeding the stable (
The presence of moist neutral conditions in the core of the AR below 2.5 km has been documented in previous studies (Ralph et al. 2005; Neiman et al. 2008a), identified by Sawyer (1956) as one of the conditions favorable for orographic enhancement of rainfall. The mean squared moist Brunt–Väisälä frequency (BVF;
Vertical distribution of water vapor transport is an important parameter as it plays an important role in orographic precipitation, which is the dominant mechanism by which the U.S. West Coast receives extreme precipitation from ARs. This study has quantified the height at which 50% and 75% of IVT is achieved, which has previously been diagnosed using a much smaller dataset. Previous studies have suggested the depth where 75% of column IVT is reached is below 2.25 km (Ralph et al. 2005) or 3 km (Ralph et al. 2017b), and in this study it has been found that in the cold sector, core, and warm sector, the mean heights are 3370 ± 930 m, 3115 ± 599 m, and 2679 ± 909 m, respectively. In the warm sector, the lower height below which 75% of column IVT is reached, along with a low-level jet at a lower altitude, are consistent with the findings in Ralph et al. (2017b). The mean PBL height for all sectors was between 650 and 835 m, with a significant positive correlation found with the lifting condensation level for all sectors, strongest in the cold sector and weakest in the core. The height below which 75% of IVT is contained was on average 2500–3000 m above the PBL height for all AR sectors, and 8%–18% of the IVT was most frequently contained within the PBL.
This study using only observations has revealed distinct characteristics across ARs when categorized into sectors based on IVT. We observe symmetry across the AR in terms of TIVT, low-level jet wind speed, and lifting condensation level, but asymmetry in other diagnostics, including height of low-level jet and height below which 75% IVT is contained. Analyzing a large sample (858 dropsondes) has allowed for examination of the variability that is apparent in the ARs, rather than simply presenting mean characteristics, which have been reported in previous studies.
The primary focus of this study is ARs, with analysis in relation to IVT, and so the detection of nearby extratropical cyclones and fronts is beyond the scope of this work. Zhang et al. (2019) found that 82% of ARs are associated with an extratropical cyclone, so although we assume most of the ARs in this work are in the vicinity of a cyclone, the position of the dropsondes and sectors relative to associated extratropical cyclones and their respective fronts introduce variability in metrics presented here.
These results are primarily representative of late winter ARs off the West Coast of the United States. Using purely observations from the CalWater and AR Recon field campaigns, we focused on the time period from 16 January to 2 March. ARs are also present during the summer months, and generally have a larger moisture content and reduced winds compared to the winter. This is also the case at low latitudes (Zhang et al. 2019), and although we exclude tropical ARs, we do not impose a latitude threshold, and dropsonde latitude spans approximately 40° in this study. The results shown therefore combine ARs that are more wind-driven with those that are more moisture driven. This analysis has only taken into account the absolute IVT of the system, therefore strong and weak ARs can be grouped together in the same sectors.
As well as the seasonal variation in AR characteristics, we note the importance of interannual variation, such as the modulation of low-level jet wind speed by El Niño. Large-scale modes of variability such as El Niño also affect the spatial distribution of ARs, with an increased frequency in the northeast Pacific during El Niño (Guan and Waliser 2015) and associated changes in landfall, increasing in the western United States in east Pacific El Niño, and in the Southwest United States in central Pacific El Niño (Kim et al. 2019). Although we have not taken into account the landfall location of the ARs in the study, the variability introduced by ENSO may impact the sampling and have an indirect impact on the results.
Atmospheric rivers are a global phenomenon, contributing to extreme precipitation in regions including East Antarctica (Gorodetskaya et al. 2014), western Europe (e.g., Lavers et al. 2012; Brands et al. 2017), the western United States (e.g., Mundhenk et al. 2016), and the Southeast United States (Lamjiri et al. 2020). Techniques used and results obtained in this study may not be appropriate for all regions. For example, Mahoney et al. (2016) found it necessary to use 500 kg m−1 s−1 for AR detection in the Southeast United States compared to the standard 250 kg m−1 s−1 threshold used here. They also found that AR events in the Southeast United States were more commonly linked to tropical cyclones than in the western United States, which would likely affect the results if this study were carried out in that location.
An important goal of AR observational campaigns is to retrieve data that will reduce forecast error and uncertainty in real time. This study has shown the value of these observations to pure research, furthering the understanding of characteristics of ARs from observations. This simple technique for identifying sectors within an AR can be applied across a variety of studies, for example in forecast diagnostics and assessing model performance. Improvement in the forecast models would allow for better prediction of landfalling ARs that bring both beneficial and damaging precipitation to the U.S. West Coast. We hope to use this study to help inform the sampling strategy of ARs, by further analyzing the sensitivity of forecasts to assimilation of dropsondes in different sectors, therefore helping to bridge the gap between observations and models. Following work also examines dropsondes in the sectors defined in this study to assess atmospheric reanalysis products, which is important as these are the closest we get to spatially homogeneous observations.
Acknowledgments
The dropsonde data examined in this study were collected during several field campaigns from 2014, 2015, 2016, 2018, 2019, and 2020 involving many scientists, engineers, air crews, project managers, program managers, and others. These include individuals from NOAA, NASA, the Air Force, and elsewhere. Without their efforts, these data would not be available for this study. Special thanks to Rich Henning from NOAA, Capt. Garrett Black and Lt. Col. Ryan Rickert from the U.S. Air Force and Holger Vömel at NCAR for providing additional information on the dropsondes. Thanks to Rich Rotunno for valuable discussion on atmospheric stability, David Lavers for comments on an early draft, and Brian Kawzenuk for assisting with data access. The authors are grateful to the three anonymous reviewers whose comments helped to clarify and improve the paper. We also acknowledge the support received from Unidata in using Metpy (Unidata 2017) functions. This work was supported by the California Department of Water Resources AR research program (Award 4600013361) and the U.S. Army Corps of Engineers Engineer Research and Development Center (Award 609 W912HZ-15-2-0019).
Data availability statement
Copernicus Climate Change Service (C3S) (2017): ERA5: Fifth generation of ECMWF atmospheric reanalyses of the global climate. Copernicus Climate Change Service Climate Data Store (CDS), November 2019. https://cds.climate.copernicus.eu/cdsapp#!/home.
REFERENCES
American Meteorological Society, 2019: Atmospheric river. Glossary of Meteorology, http://glossary.ametsoc.org/wiki/Atmospheric_river.
Brands, S., J. M. Gutiérrez, and D. San-Martín, 2017: Twentieth-century atmospheric river activity along the west coasts of Europe and North America: Algorithm formulation, reanalysis uncertainty and links to atmospheric circulation patterns. Climate Dyn., 48, 2771–2795, https://doi.org/10.1007/s00382-016-3095-6.
Browning, K., and C. Pardoe, 1973: Structure of low-level jet streams ahead of mid-latitude cold fronts. Quart. J. Roy. Meteor. Soc., 99, 619–638, https://doi.org/10.1002/qj.49709942204.
Conrick, R., and C. F. Mass, 2019: An evaluation of simulated precipitation characteristics during OLYMPEX. J. Hydrometeor., 20, 1147–1164, https://doi.org/10.1175/JHM-D-18-0144.1.
Cordeira, J. M., F. M. Ralph, A. Martin, N. Gaggini, J. R. Spackman, P. J. Neiman, J. J. Rutz, and R. Pierce, 2017: Forecasting atmospheric rivers during CalWater 2015. Bull. Amer. Meteor. Soc., 98, 449–459, https://doi.org/10.1175/BAMS-D-15-00245.1.
Corringham, T. W., F. M. Ralph, A. Gershunov, D. R. Cayan, and C. A. Talbot, 2019: Atmospheric rivers drive flood damages in the western United States. Sci. Adv., 5, eaax4631, https://doi.org/10.1126/sciadv.aax4631.
Dacre, H. F., O. Martinez-Alvarado, and C. O. Mbengue, 2019: Linking atmospheric rivers and warm conveyor belt airflows. J. Hydrometeor., 20, 1183–1196, https://doi.org/10.1175/JHM-D-18-0175.1.
Demirdjian, R., J. R. Norris, A. Martin, and F. M. Ralph, 2020: Dropsonde observations of the ageostrophy within the pre-cold-frontal low-level jet associated with atmospheric rivers. Mon. Wea. Rev., 148, 1389–1406, https://doi.org/10.1175/MWR-D-19-0248.1.
Durran, D. R., and J. B. Klemp, 1982: On the effects of moisture on the Brunt–Väisälä frequency. J. Atmos. Sci., 39, 2152–2158, https://doi.org/10.1175/1520-0469(1982)039<2152:OTEOMO>2.0.CO;2.
Eckhardt, S., A. Stohl, H. Wernli, P. James, C. Forster, and N. Spichtinger, 2004: A 15-year climatology of warm conveyor belts. J. Climate, 17, 218–237, https://doi.org/10.1175/1520-0442(2004)017<0218:AYCOWC>2.0.CO;2.
Gershunov, A., T. Shulgina, F. M. Ralph, D. A. Lavers, and J. J. Rutz, 2017: Assessing the climate-scale variability of atmospheric rivers affecting western North America. Geophys. Res. Lett., 44, 7900–7908, https://doi.org/10.1002/2017GL074175.
Gorodetskaya, I. V., M. Tsukernik, K. Claes, M. F. Ralph, W. D. Neff, and N. P. Van Lipzig, 2014: The role of atmospheric rivers in anomalous snow accumulation in East Antarctica. Geophys. Res. Lett., 41, 6199–6206, https://doi.org/10.1002/2014GL060881.
Guan, B., and D. E. Waliser, 2015: Detection of atmospheric rivers: Evaluation and application of an algorithm for global studies. J. Geophys. Res. Atmos., 120, 12 514–12 535, https://doi.org/10.1002/2015JD024257.
Guan, B., and D. E. Waliser, 2017: Atmospheric rivers in 20 year weather and climate simulations: A multimodel, global evaluation. J. Geophys. Res. Atmos., 122, 5556–5581, https://doi.org/10.1002/2016JD026174.
Guan, B., D. E. Waliser, and F. M. Ralph, 2018: An intercomparison between reanalysis and dropsonde observations of the total water vapor transport in individual atmospheric rivers. J. Hydrometeor., 19, 321–337, https://doi.org/10.1175/JHM-D-17-0114.1.
Guan, B., D. E. Waliser, and F. M. Ralph, 2020: A multimodel evaluation of the water vapor budget in atmospheric rivers. Ann. N. Y. Acad. Sci., 1472, 139–154, https://doi.org/10.1111/nyas.14368.
Hecht, C. W., and J. M. Cordeira, 2017: Characterizing the influence of atmospheric river orientation and intensity on precipitation distributions over north coastal California. Geophys. Res. Lett., 44, 9048–9058, https://doi.org/10.1002/2017GL074179.
Hersbach, H., and Coauthors, 2020: The ERA5 global reanalysis. Quart. J. Roy. Meteor. Soc., 146, 1999–2049, https://doi.org/10.1002/qj.3803.
Hodur, R. M., 1997: The Naval Research Laboratory’s Coupled Ocean/Atmosphere Mesoscale Prediction System (COAMPS). Mon. Wea. Rev., 125, 1414–1430, https://doi.org/10.1175/1520-0493(1997)125<1414:TNRLSC>2.0.CO;2.
Houze, R. A., Jr., and Coauthors, 2017: The Olympic Mountains Experiment (OLYMPEX). Bull. Amer. Meteor. Soc., 98, 2167–2188, https://doi.org/10.1175/BAMS-D-16-0182.1.
Hughes, M., K. M. Mahoney, P. J. Neiman, B. J. Moore, M. Alexander, and F. M. Ralph, 2014: The landfall and inland penetration of a flood-producing atmospheric river in Arizona. Part II: Sensitivity of modeled precipitation to terrain height and atmospheric river orientation. J. Hydrometeor., 15, 1954–1974, https://doi.org/10.1175/JHM-D-13-0176.1.
Kim, H.-M., Y. Zhou, and M. A. Alexander, 2019: Changes in atmospheric rivers and moisture transport over the Northeast Pacific and western North America in response to ENSO diversity. Climate Dyn., 52, 7375–7388, https://doi.org/10.1007/s00382-017-3598-9.
Knippertz, P., H. Wernli, H. Binder, M. Böttcher, H. Joos, E. Madonna, G. Pante, and M. Sprenger, 2018: The relationship between warm conveyor belts, tropical moisture exports and atmospheric rivers. Geophysical Research Abstracts, Vol. 20, Abstract EGU2018-4362, https://meetingorganizer.copernicus.org/EGU2018/EGU2018-4362.pdf.
Lackmann, G. M., 2002: Cold-frontal potential vorticity maxima, the low-level jet, and moisture transport in extratropical cyclones. Mon. Wea. Rev., 130, 59–74, https://doi.org/10.1175/1520-0493(2002)130<0059:CFPVMT>2.0.CO;2.
Lamjiri, M. A., F. M. Ralph, and M. D. Dettinger, 2020: Recent changes in U.S. extreme 3-day precipitation using the R-CAT scale. J. Hydrometeor., 21, 1207–1221, https://doi.org/10.1175/JHM-D-19-0171.1.
Lavers, D. A., G. Villarini, R. P. Allan, E. F. Wood, and A. J. Wade, 2012: The detection of atmospheric rivers in atmospheric reanalyses and their links to British winter floods and the large-scale climatic circulation. J. Geophys. Res., 117, D20106, https://doi.org/10.1029/2012JD018027.
Lavers, D. A., M. J. Rodwell, D. S. Richardson, F. M. Ralph, J. D. Doyle, C. A. Reynolds, V. Tallapragada, and F. Pappenberger, 2018: The gauging and modeling of rivers in the sky. Geophys. Res. Lett., 45, 7828–7834, https://doi.org/10.1029/2018GL079019.
Lavers, D. A., A. Beljaars, D. S. Richardson, M. J. Rodwell, and F. Pappenberger, 2019: A forecast evaluation of planetary boundary layer height over the ocean. J. Geophys. Res. Atmos., 124, 4975–4984, https://doi.org/10.1029/2019JD030454.
Leung, L. R., and Y. Qian, 2009: Atmospheric rivers induced heavy precipitation and flooding in the western U.S. simulated by the WRF regional climate model. Geophys. Res. Lett., 36, L03820, https://doi.org/10.1029/2008GL036445.
Mahoney, K., and Coauthors, 2016: Understanding the role of atmospheric rivers in heavy precipitation in the southeast United States. Mon. Wea. Rev., 144, 1617–1632, https://doi.org/10.1175/MWR-D-15-0279.1.
Miglietta, M., and R. Rotunno, 2005: Simulations of moist nearly neutral flow over a ridge. J. Atmos. Sci., 62, 1410–1427, https://doi.org/10.1175/JAS3410.1.
Morales, A., H. Morrison, and D. J. Posselt, 2018: Orographic precipitation response to microphysical parameter perturbations for idealized moist nearly neutral flow. J. Atmos. Sci., 75, 1933–1953, https://doi.org/10.1175/JAS-D-17-0389.1.
Mundhenk, B. D., E. A. Barnes, and E. D. Maloney, 2016: All-season climatology and variability of atmospheric river frequencies over the North Pacific. J. Climate, 29, 4885–4903, https://doi.org/10.1175/JCLI-D-15-0655.1.
Neiman, P. J., F. M. Ralph, A. White, D. Kingsmill, and P. Persson, 2002: The statistical relationship between upslope flow and rainfall in California’s coastal mountains: Observations during CALJET. Mon. Wea. Rev., 130, 1468–1492, https://doi.org/10.1175/1520-0493(2002)130<1468:TSRBUF>2.0.CO;2.
Neiman, P. J., F. M. Ralph, G. A. Wick, Y.-H. Kuo, T.-K. Wee, Z. Ma, G. H. Taylor, and M. D. Dettinger, 2008a: Diagnosis of an intense atmospheric river impacting the Pacific Northwest: Storm summary and offshore vertical structure observed with COSMIC satellite retrievals. Mon. Wea. Rev., 136, 4398–4420, https://doi.org/10.1175/2008MWR2550.1.
Neiman, P. J., F. M. Ralph, G. A. Wick, J. D. Lundquist, and M. D. Dettinger, 2008b: Meteorological characteristics and overland precipitation impacts of atmospheric rivers affecting the West Coast of North America based on eight years of SSM/I satellite observations. J. Hydrometeor., 9, 22–47, https://doi.org/10.1175/2007JHM855.1.
Neiman, P. J., M. Hughes, B. J. Moore, F. M. Ralph, and E. M. Sukovich, 2013a: Sierra barrier jets, atmospheric rivers, and precipitation characteristics in Northern California: A composite perspective based on a network of wind profilers. Mon. Wea. Rev., 141, 4211–4233, https://doi.org/10.1175/MWR-D-13-00112.1.
Neiman, P. J., F. M. Ralph, B. J. Moore, M. Hughes, K. M. Mahoney, J. M. Cordeira, and M. D. Dettinger, 2013b: The landfall and inland penetration of a flood-producing atmospheric river in Arizona. Part I: Observed synoptic-scale, orographic, and hydrometeorological characteristics. J. Hydrometeor., 14, 460–484, https://doi.org/10.1175/JHM-D-12-0101.1.
Oakley, N. S., J. T. Lancaster, M. L. Kaplan, and F. M. Ralph, 2017: Synoptic conditions associated with cool season post-fire debris flows in the Transverse Ranges of southern California. Nat. Hazards, 88, 327–354, https://doi.org/10.1007/s11069-017-2867-6.
OFCM, 2019: National Winter Season Operations Plan (NWSOP). Office of the Federal Coordinator for Meteorological Services and Supporting Research, 84 pp., www.ofcm.gov/publications/nwsop/2019_nwsop.pdf.
Ralph, F. M., and M. Dettinger, 2012: Historical and national perspectives on extreme West Coast precipitation associated with atmospheric rivers during December 2010. Bull. Amer. Meteor. Soc., 93, 783–790, https://doi.org/10.1175/BAMS-D-11-00188.1.
Ralph, F. M., P. J. Neiman, and G. A. Wick, 2004: Satellite and CALJET aircraft observations of atmospheric rivers over the eastern North Pacific Ocean during the winter of 1997/98. Mon. Wea. Rev., 132, 1721–1745, https://doi.org/10.1175/1520-0493(2004)132<1721:SACAOO>2.0.CO;2.
Ralph, F. M., P. J. Neiman, and R. Rotunno, 2005: Dropsonde observations in low-level jets over the northeastern Pacific Ocean from CALJET-1998 and PACJET-2001: Mean vertical-profile and atmospheric-river characteristics. Mon. Wea. Rev., 133, 889–910, https://doi.org/10.1175/MWR2896.1.
Ralph, F. M., P. J. Neiman, G. A. Wick, S. I. Gutman, M. D. Dettinger, D. R. Cayan, and A. B. White, 2006: Flooding on California’s Russian River: Role of atmospheric rivers. Geophys. Res. Lett., 33, L13801, https://doi.org/10.1029/2006GL026689.
Ralph, F. M., and Coauthors, 2014: A vision for future observations for western US extreme precipitation and flooding. J. Contemp. Water Res. Educ., 153, 16–32, https://doi.org/10.1111/j.1936-704X.2014.03176.x.
Ralph, F. M., and Coauthors, 2016: CalWater field studies designed to quantify the roles of atmospheric rivers and aerosols in modulating U.S. West Coast precipitation in a changing climate. Bull. Amer. Meteor. Soc., 97, 1209–1228, https://doi.org/10.1175/BAMS-D-14-00043.1.
Ralph, F. M., and Coauthors, 2017a: Atmospheric rivers emerge as a global science and applications focus. Bull. Amer. Meteor. Soc., 98, 1969–1973, https://doi.org/10.1175/BAMS-D-16-0262.1.
Ralph, F. M., and Coauthors, 2017b: Dropsonde observations of total integrated water vapor transport within North Pacific atmospheric rivers. J. Hydrometeor., 18, 2577–2596, https://doi.org/10.1175/JHM-D-17-0036.1.
Ralph, F. M., J. J. Rutz, J. M. Cordeira, M. Dettinger, M. Anderson, D. Reynolds, L. J. Schick, and C. Smallcomb, 2019: A scale to characterize the strength and impacts of atmospheric rivers. Bull. Amer. Meteor. Soc., 100, 269–289, https://doi.org/10.1175/BAMS-D-18-0023.1.
Ralph, F. M., and Coauthors, 2020: West Coast forecast challenges and development of atmospheric river reconnaissance. Bull. Amer. Meteor. Soc., 101, E1357–E1377, https://doi.org/10.1175/BAMS-D-19-0183.1.
Reynolds, C. A., J. D. Doyle, F. M. Ralph, and R. Demirdjian, 2019: Adjoint sensitivity of North Pacific atmospheric river forecasts. Mon. Wea. Rev., 147, 1871–1897, https://doi.org/10.1175/MWR-D-18-0347.1.
Rutz, J. J., W. J. Steenburgh, and F. M. Ralph, 2014: Climatological characteristics of atmospheric rivers and their inland penetration over the western United States. Mon. Wea. Rev., 142, 905–921, https://doi.org/10.1175/MWR-D-13-00168.1.
Sawyer, J., 1956: The physical and dynamical problems of orographic rain. Weather, 11, 375–381, https://doi.org/10.1002/j.1477-8696.1956.tb00264.x.
Seidel, D. J., Y. Zhang, A. Beljaars, J.-C. Golaz, A. R. Jacobson, and B. Medeiros, 2012: Climatology of the planetary boundary layer over the continental United States and Europe. J. Geophys. Res., 117, D17106, https://doi.org/10.1029/2012JD018143.
Smith, B. L., S. E. Yuter, P. J. Neiman, and D. Kingsmill, 2010: Water vapor fluxes and orographic precipitation over northern California associated with a landfalling atmospheric river. Mon. Wea. Rev., 138, 74–100, https://doi.org/10.1175/2009MWR2939.1.
Torn, R. D., and G. J. Hakim, 2008: Ensemble-based sensitivity analysis. Mon. Wea. Rev., 136, 663–677, https://doi.org/10.1175/2007MWR2132.1.
Unidata, 2017: Metpy: A Python package for meteorological data. UCAR/Unidata, accessed 1 January 2020, https://doi.org/10.5065/D6WW7G29.
Vaisala, 2017: Dropsonde RD94. RD94 datasheet, Vaisala, 2 pp., https://www.vaisala.com/sites/default/files/documents/RD94-Datasheet-B210936EN-B.pdf.
Vaisala, 2018: Dropsonde RD41. RD41 datasheet, Vaisala, 2 pp., https://www.vaisala.com/sites/default/files/documents/RD41-Datasheet-B211706EN.pdf.
Vallis, G. K., 2017: Atmospheric and Oceanic Fluid Dynamics: Fundamentals and Large-scale Circulation. Cambridge University Press, 965 pp.
Wallace, J. M., and P. V. Hobbs, 2006: Atmospheric Science: An Introductory Survey. Vol. 92, Elsevier, 504 pp.
Wick, G. A., P. J. Neiman, F. M. Ralph, and T. M. Hamill, 2013: Evaluation of forecasts of the water vapor signature of atmospheric rivers in operational numerical weather prediction models. Wea. Forecasting, 28, 1337–1352, https://doi.org/10.1175/WAF-D-13-00025.1.
Yu, C.-K., and B. F. Smull, 2000: Airborne Doppler observations of a landfalling cold front upstream of steep coastal orography. Mon. Wea. Rev., 128, 1577–1603, https://doi.org/10.1175/1520-0493(2000)128<1577:ADOOAL>2.0.CO;2.
Zhang, Z., F. M. Ralph, and M. Zheng, 2019: The relationship between extratropical cyclone strength and atmospheric river intensity and position. Geophys. Res. Lett., 46, 1814–1823, https://doi.org/10.1029/2018GL079071.