1. Introduction
Aeolian dust is one of the most pervasive aerosols in Earth’s atmosphere (Huneeus et al. 2011). Dust alters the planet’s radiative budget and hydrological cycles via aerosol direct and indirect effects (Choobari et al. 2014) and affects nutrient cycling in the marine and terrestrial ecosystems where dust emission and deposition occurs (Field et al. 2010). As such, there is a need to understand how planetary climate change has—and will continue to—influence the processes of dust emission, transport, and deposition, the so-called dust cycle (Shao et al. 2011), as well as to understand how those forced changes in the dust cycle feed back onto Earth’s climate (Kok et al. 2018). However, studies examining the representation of dust in model output from the fifth and sixth Climate Model Intercomparison Projects have identified model biases in the dust mean state, poor reproduction of historical dust variability, and insufficient sensitivity of dust emission to changes in surface conditions (Pu and Ginoux 2018; Zhao et al. 2022), casting doubt on our ability to model future dust.
Improving understanding of the physical processes leading to dust emission and transport can lead to advances in the representation of dust in models. Although there is a growing body of knowledge of the meteorological processes underlying dust storms (Knippertz 2014), there remains a dearth of representative in situ observations in dust emitting regions, which is not entirely surprising given that most dust outbreaks occur in sparsely populated regions (Prospero et al. 2002) where challenges associated with access can be significant (e.g., Giles 2005). This study aims to add to understanding of the meteorological processes affecting dust storms by examining measurements made during three dust outbreaks in a region of southeastern California, with a specific focus on the role of complex terrain in shaping the characteristics of the high winds and lofted dust. Previous studies have identified several processes associated with orographically forced flow that result in high winds and dust lofting, including gap flow (Evan et al. 2016; Jiang et al. 2009; Todd et al. 2008), downslope winds due to orographic precipitation and latent cooling of air (Knippertz et al. 2007; Evan et al. 2022c), generic foehn events (Gläser et al. 2012; Evan 2019), and leeside rotor circulations (Grubišić and Billings 2007; Pokharel et al. 2017). Here we focus on the role of trapped lee waves in generating dust outbreaks.
Trapped lee waves are a class of orographically forced waves (i.e., generated by air flowing over a mountain range) for which the waves are trapped in the lower atmosphere, rather than propagating upward through the troposphere (Nappo 2013), propagating laterally well beyond the location of wave generation (Durran 2003). Vertical variations in stability and shear in the upstream flow (i.e., upwind of the mountain range) give rise to trapped waves (Scorer 1949), and temporal changes in these properties result in nonstationary waves (Ralph et al. 1997). Trapped waves can give rise to rotors in the downslope flow, in which rapid vertical ascent in the upward branch of a wave can produce flow separation at the surface and reversed surface winds under the wave crest (Doyle and Durran 2002), and modify (both accelerate and decelerate) surface wind speeds far beyond the wave source (Durran 1986).
While there is a rich history of scholarly work on the topic of trapped lee waves (cf. Smith 2019), to the best of our knowledge studies connecting trapped waves to dust emission and transport have been limited to the Owens valley, and more strongly focused on the dynamics of the leeside circulation than the characteristics of the subsequent dust storms (Grubišić et al. 2008; De Wekker and Mayor 2009; Jiang et al. 2011; Strauss et al. 2016). Additionally, Owens valley is narrow and consequently waves forming in the lee of the eastern Sierra are distinct from trapped lee waves that are able to propagate long distances downwind of the region of wave generation. Given the ubiquity of trapped lee waves in Earth’s atmosphere it is at least plausible that these orographically forced phenomena are responsible for a nonnegligible fraction of the global dust uplift (e.g., downwind of the Atlas or Andes Mountains).
Our area of interest is the Salton basin, a sub-sea-level terminal basin located at the northwestern corner of the Sonoran Desert that is part of the greater Salton trough, a northwest–southeast oriented rift valley along the San Andreas Fault (Fig. 1). At the lowest elevations of the basin lies the Salton Sea, an endorheic body of water having an average surface height of −72.7 m MSL in 2021 (dashboard.waterdata.usgs.gov, accessed on 24 March 2022). Dust storms are a frequent occurrence in this region (Evan 2019), which is due in part to the prevalence of erodible soils (Buck et al. 2011; Sweeney et al. 2011). The Salton Sea was accidentally created in 1905 during an attempt to irrigate the southern portion of the Salton trough (the Imperial Valley), but more recently the volume of the sea has been declining due to a 2003 water transfer agreement that resulted in diversion of water from the sea. Consequently, the size of the Salton Sea is rapidly declining (Poudel et al. 2021).
Terrain of the region of interest. (a) A true color image acquired from MODIS Aqua at 2100 UTC 2 Mar 2021. (b) An elevation map of the same region. The approximate shoreline of the Salton Sea during March 2021 is indicated by the gray contour. The locations of the field and the NKX radiosonde sites are indicated by the white circles and triangles, respectively, in both panels. The desert that lies immediately west of the field site is the source region for the airborne dust measured at the site.
Citation: Journal of the Atmospheric Sciences 80, 3; 10.1175/JAS-D-22-0128.1
Playa sources represent a significant fraction of all dust emission associated with human activity (Ginoux et al. 2012), and the drying of bodies of water in arid regions increases the incidence and intensity of dust storms there (Zucca et al. 2021). A simulation of a single dust event in the Salton basin estimated an approximately 10% increase in dust burden with a nearly 40% growth in the playa surface (Parajuli and Zender 2018), which is significant given that the playa is surrounded by desert dust sources that are vastly larger in spatial extent. Other work has shown adverse health effects from exposure to dust emitted from the playa (Burr et al. 2021; Biddle et al. 2021), which contains anthropogenic trace metals (Frie et al. 2019). As such, improving understanding of the meteorology underlying dust events in this region is useful in terms of understanding the changing dust burden and the associated human health impacts.
The remainder of this article is organized as follows. In section 2 we describe the observational data and model output used in the study. In section 3 we examine the meteorological and physical aspects of three dust storms via measurements and model output. In section 4 we discuss general characteristics of dust storms generated by trapped lee waves. In section 5 we summarize the work presented here, note the broader implications of the findings, and suggest additional observations and modeling studies to address remaining questions.
2. Observations and model
We start by describing the region of interest (Fig. 1). The Salton basin is an arid endorheic basin that typically receives less than 100 mm of precipitation each year (Stephen and Gorsline 1975; NCEI 2022). The morphology of the area includes alluvial fans, sand and sand dunes, dry washes, paleo lakebed, and rock and vegetated surfaces (IID 2016). Within the basin, the Salton Sea is a spatially large yet shallow inland body of water. Agriculture land is found immediately to the north and south of the Salton Sea, whereas the Anza Desert, from which many dust storms in the area originate, lies immediately to its west (Fig. 1a). The basin is bounded to the west by the Peninsular Range, to the north by the San Bernardino Mountains, and to the east by the Transverse Range, while the topography gradually slopes upward to the south before dropping into the Colorado River delta (Fig. 1b).
a. Field site and in situ observations
Much of the observational data presented here were collected from a field site located near the current western coastline of the Salton Sea, at approximately 33.2°N, −115.9°E (Fig. 1). The site is adjacent to a large citrus and date palm farm, which provides physical security for the station and allows for access to a stable source of power for instrumentation and telemetry. The landscape immediately surrounding the site is characterized by narrow dry washes and cobbles distributed over silt-dominated paleo lakebed with sparse shrub vegetation.
An AERONET CIMEL Electronique sun–sky photometer is located at the site, which is used to measure sun collimated direct beam irradiance and directional sky radiance at eight spectral bands centered on 1020, 870, 675, 440, 936, 500, 380, and 340 nm (Holben et al. 1998). The instrument base is mounted approximately 2 m above ground level. Direct solar irradiance measurements are made at 5-min intervals. Here we utilize data from the AERONET level 1.5 products processed by the version 3 AERONET algorithm, which provides fully automatic cloud screening and instrument anomaly quality controls in near–real time (Giles et al. 2019). We include dusty observations that were erroneously classified as cloud contaminated using the restoring algorithm described in Evan et al. (2022a).
Located at the field site is a Vaisala CL51 ceilometer, which is a single lens lidar system that makes continuous profiles of attenuated backscatter at a nominal wavelength of 910 nm and up to heights of 15 km. The CL51 range corrected backscatter profiles used here are generated at a 36-s temporal resolution and a 10-m vertical resolution. In addition to cloud detection, ceilometers, including the CL51, have shown to be useful in the detection of aerosol layers in the lower troposphere (Münkel et al. 2007; Wiegner et al. 2014; Jin et al. 2015; Marcos et al. 2018; Yang et al. 2020). The Vaisala processing software for the CL51 measurements, BLView, produces retrievals of vertical profiles of extinction σ and optical depth τ from the backscatter profiles for the clear-sky atmosphere below 5 km. Although details regarding the retrieval process used in BLView are not publicly available, we are able to approximately reproduce the extinction profile retrievals using the methods described in (Fernald 1984), as discussed in Evan et al. (2022c). We calibrate the 910-nm aerosol optical depth (AOD) retrieved from the CL51, which is obtained by integrating the retrieved extinction profiles in the vertical dimension to an equivalent 500-nm value by comparing values of 500-nm AOD from AERONET to the 910-nm AOD retrieved from the CL51, following the methods in Evan et al. (2022a).
At this site also sits a cabled Vantage Pro2 Davis Met Station, which has a suite of sensors including temperature and humidity sensors under a passive radiation shield, a wind anemometer, a barometer, and rainfall measurements. Data are logged at a 1-min interval. The site anemometer sits approximately 2 m above ground level. The 2-m wind speed and gust measurements were calibrated to an equivalent 10-m wind speed value by multiplying the 2-m values by a factor of 1.37, which was empirically derived via comparison to an adjacent 10-m mounted anemometer (Evan et al. 2022a). We note that at present only hourly averaged values are available from the 10-m anemometer, which is managed by the local water and power utility, Imperial Irrigation District.
Vertical profiles of temperature, humidity, pressure, and wind are obtained from Vaisala RS-41 sondes launched at the site on 9 March 2021 (at 1203, 1505, 1803, 1934, 2102, 2234, and 2359 UTC), 15 March 2021 (at 2114 and 2320 UTC), and 15 February 2022 (at 2105, 2242, and 2340 UTC). Last, all heights from soundings and the CL51 are referenced to ground level of the station, which sits approximately 32 m below mean sea level. Radiosonde, meteorological station, and ceilometer profiles made at the field site are permanently archived and publicly available (Evan et al. 2022b).
b. Other data
In addition to the measurements made at the field site we utilize surface meteorological and PM10 measurements made from stations around the Salton Sea. We examine measurements of PM10 since in this area elevated values of PM10 are an unambiguous indication of the presence of suspended dust, whereas, for example, PM2.5 may indicate the presence of dust as well as other local sources of particulates. The meteorological and PM10 data were accessed via the MesoWest network (Horel et al. 2002) and the California Air Resources Board Air Quality and Meteorological Information System. We also utilize imagery from a 360° Roundshot web camera that is located 28 km west of the field site at an elevation of 300 m AGL, which are available at approximately 10-min intervals during daytime hours. We incorporate into our analysis satellite imagery from the Moderate Resolution Imaging Spectroradiometer (MODIS) flying on board the Aqua satellite, which were generated from the NASA Earth Observing System Data and Information System (EOSDIS) Worldview application. We also generated imagery from radiance measurements made by the Advanced Baseline Imager (ABI) flying on board GOES-17. These data were accessed from the NOAA Comprehensive Large Array-data Stewardship System. We examine measurements collected from radiosondes launched from the NKX sounding station, which is near the coastline (white triangle, Fig. 1), where radiosondes are launched twice daily at 0000 and 2400 UTC. Three-hourly output from the North American Regional Reanalysis (NARR), which is provided on 29 vertical layers at a 32-km horizontal resolution, is used to examine the synoptic environment associated with the dust outbreaks studied here (Mesinger et al. 2006).
c. WRF Model
Numerical simulations of the meteorology underlying the dust cases examined here were made using the Advanced Research version of the Weather Research and Forecasting (WRF) Model (Skamarock et al. 2019) version 4.3. The model was run using three-domain, nested two-way interactive grid with horizontal resolutions of 15, 5, and 1 km (Fig. 2). The model was initialized using data from the Global Forecast System (GFS) output (NCEP 2013) at 0600 UTC on 8 March 2021, 14 March 2022, and 14 February 2022, and was integrated forward for the subsequent 72 h for each case with the lateral boundaries of the outermost domain continuously forced by the GFS output. WRF Model output shown here is from the innermost domain.
Domains for the nested WRF simulations. Plotted in blue are the horizontal extents of the nested domains utilized in the WRF simulations. The horizontal resolutions of the outermost to innermost domains are 15, 5, and 1 km, respectively.
Citation: Journal of the Atmospheric Sciences 80, 3; 10.1175/JAS-D-22-0128.1
The model top is at 10 hPa and 51 sigma vertical levels are employed, with the highest vertical resolution found in the lower troposphere. Approximately 7 half-sigma levels are found in the lowest kilometer AGL, with the first level at a height of 27 m AGL. The output shown here is from simulations using the Mellor–Yamada–Janjić (Janjić 1994) planetary boundary layer scheme, which, when compared to other boundary layer schemes, was found to best reproduce in situ observations, particularly the surface wind speeds, in the region (Evan et al. 2022c). The model physics parameterizations used in this study are shown in Table 1. Comparisons of WRF output to surface wind measurements at the field site and radiosondes launched during the dust outbreaks considered here can be found in online supplemental Figs. S1–S13.
Physics schemes employed in the WRF simulations.
We also conduct simulations using the WRF-Chem model (Grell et al. 2005; Fast et al. 2006; Peckham et al. 1991), employing the GOCART aerosol scheme without ozone chemistry (Chin et al. 2000; Ginoux et al. 2001) and the Air Force Weather Agency dust emission scheme (AFWA; LeGrand et al. 2019), with other model parameterizations, setup, and forcing identical to that described for the WRF simulations (Table 1). The AFWA emissions scheme, which uses a modified version of the saltation-based dust emission function of Marticorena and Bergametti (1995), is one of several available by default in current versions of WRF-Chem. This scheme represents an update to the earlier GOCART-WRF emissions scheme, incorporating separately modeled saltation processes driving subsequent dust emissions rather than the single-step parameterization used previously. Since its addition to WRF-Chem, the AFWA scheme has been used and evaluated in dust modeling research and case studies around the world (e.g., Yuan et al. 2019; Kim et al. 2021; Miller et al. 2021). In the model dust in the size range of 0.2–20 μm is simulated in five bins.
When comparing the model simulated dust to aerosol measurements at the field site and surface PM10 measurements at a number of locations around the Salton Sea we found that the model produced too much dust at weak wind speeds and too small an increase in dust as the wind speed increased. We also found via comparison to surface PM10 measurements that dust surface concentrations were biased low in the region of the research site, and biased high to the north and south of the site (not shown). These biases persisted across two different soil erodibility input maps, including the default GOCART topographic erodibility dataset of Ginoux et al. (2001), as well as the more recent dataset of Parajuli and Zender (2017). Based on their consistency across erodibility map inputs, we suspect that these biases are related to other surface property inputs, such as soil and land surface cover type.
Due to concerns over the representation of dust emission in the model we only utilize output from a WRF-Chem simulation of the dust outbreak on 15 March 2021 in order to examine the general relationship between trapped lee wave phase and the vertical and horizontal distribution of dust (section 4). We leave improvement of the representation of modeled dust in this region for future work.
d. Salton Sea extent
The extent of the Salton Sea was estimated using MODIS Aqua visible satellite imagery from 2 March 2021 (Fig. 1a). To estimate the shoreline, we applied an arbitrary threshold to the reflectances of each of the three image color channels (i.e., red, green, blue) in order to distinguish the dark Salton Sea against the bright desert surface, manually excluding any pixels that were dark enough to pass this threshold test from the vegetated croplands to the south of the sea. We then used these data to define the shoreline of the sea (gray contour, Fig. 1b). The shoreline estimate is used as a visual aid in several figures found throughout this manuscript.
3. Characteristics of the dust storms
Here we consider three dust outbreaks within the Salton basin: 9 and 15 March 2021, and 15 February 2022. When convenient we only refer to these cases using their respective months and days. These events were chosen because of the similarities in their meteorological aspects and dust characteristics, and the availability of radiosonde measurements made at the field site.
a. Synoptic situation
We first describe the synoptic environments for the three dust events considered here. For all cases streamlines and heights of the 500 hPa pressure surfaces from NARR show an upper-level low displaced to the northwest of the region of interest, with the lows’ centers of action approximately located at 45°, −130°E on 9 March 2021 (Fig. 3a), and 40°N, −120°E on 15 March 2021 and 15 February 2022 (Figs. 3b,c, respectively). For each case the elevated lows direct westerly flow across the region of interest (black squares), with all exhibiting tightly packed height contours and cross barrier (i.e., westerly) wind speeds greater than 20 m s−1. We note that that for the 9 March case the westerly flow is driven by both the broad low located to the north over the Pacific and an anticyclone located to the southeast (anticyclone not seen in Fig. 3a). Sea level pressure contours for these three cases show surface low pressure centers north of the Salton Sea and near exit regions of the upper-level jets, with trailing cold fronts pushing through the Salton basin at approximately 1800, 2100, and 1500 UTC (Figs. 3d–f, respectively). Temperatures and vector winds at 850 hPa imply low-level northwesterly cold-air advection behind the fronts and westerly flow directed at the coastline and over the Salton Sea. The synoptic situations for these cases are similar to that described for dust outbreaks occurring on 22 February 2020 (Evan et al. 2022c) and 14 March 2018 (Evan 2019).
Synoptic situations immediately preceding the dust outbreaks: (a),(d) 1800 UTC 9 Mar 2021, (b),(e) 2100 UTC 15 Mar 2021, and (c),(e) 1500 UTC 15 Feb 2022. (a)–(c) Maps of NARR 500-hPa wind speeds (shading), streamlines (black), and heights (yellow contours). Heights of the 500-hPa pressure surfaces are represented by the yellow contours at intervals of 5 dm, with the thick contour representing the 560-dm surface. The black box indicates the area shown in Fig. 1. (d)–(f) Maps of 850-hPa temperature (shading), sea level pressure (black), and vector winds (arrows). Cold fronts (blue) and surface lows (boxed “L”) locations are based on NOAA Weather Prediction Center surface analysis. Sea level pressure contours are hPa greater than 1000 hPa. The magenta shading represents the location of the Salton Sea. The horizontal extents of the maps in (a)–(c) and (d)–(f) are not identical.
Citation: Journal of the Atmospheric Sciences 80, 3; 10.1175/JAS-D-22-0128.1
The characteristics of these mature cyclone wave and frontal systems (Fig. 3) generate unique conditions that are favorable for trapping waves, including low-level cold air-advection below a westerly jet streak. For the 9 March case (Figs. 3a,d) the upper-level trough is open and exhibits a slight negative tilt. Vertical profiles of potential temperature θ and wind speed and direction made from radiosondes launched from the NXK sounding station (see location in Fig. 1) at 1200 UTC on this day show a 5°C increase in θ in the 725–775-hPa layer (Fig. 4a), which is within a deeper layer (700–800 hPa) of backing winds (Fig. 4c), implying low-level cold-air advection and cold-frontal passage. The sounding made 12 h later on this day, and after the surface front had passed over the Salton Sea region, shows lifting of the isentropic surfaces from 500 to 800 hPa (Fig. 4a), indicating a deeper layer of cold air. The 30 m s−1 increase in wind speed from 300 to 500 hPa reflects displacement of the associated jet streak over the region (Fig. 4b). The westerly flow throughout much of the troposphere in the later sounding (Fig. 4c) reflects the strongly zonal nature of the jet at this latitude, which results in part from the continued deepening and southward migration of the low at 45°N, −130°E (not shown).
Measurements from soundings made at the NKX station (see location in Fig. 1). Vertical profiles of (a),(d),(g) potential temperature θ, (b),(e),(h) wind speed, and (c),(f),(i) wind direction collected from radiosondes launched at 1200 (light blue) and 2400 (rust) UTC (a)–(c) 9 Mar 2021, (d)–(f) 15 Mar 2021, and (g)–(i) 15 Feb 2022.
Citation: Journal of the Atmospheric Sciences 80, 3; 10.1175/JAS-D-22-0128.1
The low in the 15 March case (Figs. 3b,e) is better developed than that for 9 March, exhibiting a neutrally tilted trough digging down the western U.S. coastline. The NXK soundings from this day (Figs. 4d–f) are similar to those from 9 March in several ways, including a 6°C increase in θ at 850 hPa and a layer of backing winds from 800 to 850 hPa. The sounding made 12 h later shows lifting of the θ inversion layer to 775 hPa and the layer of backing winds to 775–825 hPa heights. The later sounding also suggests warm-air advection in the 750–600 hPa layer, as evidenced by the veering flow from 750 to 600 hPa and similarity in θ at those heights over the 12-h time period. Similar to the 9 March case is the presence of a westerly jet with maximum wind speeds at 400–300 hPa.
Last, the February 15 positively tilted short wave trough (Fig. 3c) was a fast-moving system and neither NXK sounding for this day exhibits clear signs of cold-frontal passage (Figs. 4g–i). The measurements indicate a large 8°C increase in θ at 870 hPa in the 1200 UTC sounding, which lifts to approximately 775 hPa 12 h later, with cooling throughout the 400–900-hPa heights during this time period. The latter sounding also shows an approximately 15 m s−1 increase in wind speed during this period in the 500–400-hPa layer. Noting that the veering flow below 850 hPa may be the result of surface friction rather than indicating warm-air advection, both radiosondes suggest a deep layer of positive zonal flow.
As we discuss in section 4, these profiles all exhibit characteristics favorable to the generation of trapped lee waves, including low-level cold-air advection, with warm-air advection aloft in the 15 February case, strongly zonal (i.e., cross-barrier) flow, and positively sheared winds, especially above the heights of the mountain ridges, which in Fig. 4 is approximately located in the 850–800-hPa range.
Within the Salton basin the passage of these three frontal systems generated a similar response in the surface meteorological conditions. During each the 30-min-averaged surface wind speeds Us and gusts measured at the field site exceeded 10 and 20 m s−1, respectively, and were westerly in direction (Fig. 5). The persistently westerly flow during the dust outbreaks is in contrast to the typical patterns of wind speed and direction in the basin, which can be characterized as a thermally driven daytime upslope (easterly) and downslope (westerly) circulation forced by the mountains that lie to the west of the site (e.g., 6–8 March, 12–14 March, Fig. 5a).
Time series of surface meteorological measurements made from the field site during (a) March 2021 and (b) February 2022. Plotted are 30-min-averaged values of surface wind speed Us (blue solid line), wind gust speed (blue dotted line), and wind direction (black solid line), with the value of direction indicated by the leftmost vertical axis. The gray shaded regions indicate 24-h periods commencing at 1200 UTC 9 and 15 Mar 2021, and 15 Feb 2022, during which the dust outbreaks occurred.
Citation: Journal of the Atmospheric Sciences 80, 3; 10.1175/JAS-D-22-0128.1
b. Observations of dust
We next consider the spatial and temporal variability of the dust generated by the high winds present over the Salton basin. To simultaneously visualize wind speed and direction and PM10 we generated modified versions of wind roses. For each, the station physical location is at the center point of the rose. Concentric circles indicate wind speed ranges, where the area from the center point to the first concentric circle represents wind speeds in the range of 0–4 m s−1, the area from the first to the second circles represents wind speeds in the range of 4–8 m s−1, and so on in increments of 4 m s−1. The radial divisions represent wind direction. Shading refers to the natural logarithm of the maximum hourly PM10 measured for a given wind speed and direction range, where ln PM10 values ≥ 5 are above the U.S. EPA 24-h air quality standard of 150 μg m3. The data displayed in Fig. 6 correspond to the time periods highlighted in gray in Fig. 5. Hourly averaged PM10 and wind speed measurements are used to generate these plots.
Modified wind roses indicating peak concentrations of PM10 during the three dust outbreaks. Shown in each map are roses (see text for description) made from measurements collected during the dust outbreaks on (a) 9 Mar and (b) 15 Mar 2021, and (c) 15 Feb 2022. Gray shaded rose sections in (b) indicate wind speeds and directions for which corresponding PM10 measurements were missing. The gray contours represent surface elevations at intervals of 250 m, and the thick black line represents an estimate of the Salton Sea shoreline in March 2021. The location of the field site is indicated by the red circular marker.
Citation: Journal of the Atmospheric Sciences 80, 3; 10.1175/JAS-D-22-0128.1
For all of three cases PM10 values exceeding 150 μg m−3 (ln PM10 ≥ 5) were observed for at least five of the seven stations, with PM10 exceeding 150 μg m−3 at all stations during the 15 March case (Fig. 6b). At the northernmost station the strongest wind speed and PM10 values occur during northwesterly winds, likely due to flow channeling through Banning Pass (Ryerson et al. 2013), which sits at the northern terminus of the Salton trough (Fig. 1b). Farther to the south the strongest wind speeds and PM10 values correspond to increasingly westerly flow, which reflects the widening of the basin and the proximity of the Anza Desert, which lies to the west of the Salton Sea and is upwind of the field site (Fig. 1a). Based on the prevalence of measurements for which ln PM10 ≥ 5, the 15 February event exhibited the largest number of high surface dust concentrations (Fig. 6c), and the 9 March case exhibited the lowest surface dust concentrations (Fig. 6a).
Retrievals of aerosol optical depth τ from the CIMEL sun photometer show maximum aerosol optical depths of approximately 0.4 on 9 and 15 March, and 0.45 on 15 February (Figs. 7a–c, respectively). The number of CIMEL τ retrievals is related to the presence of daytime clear-sky conditions; cloud cover was present over the site prior to 1800 UTC 9 March, and there was intermittent cloud cover throughout the 15 March event, whereas the sky was clear on 15 February (see animations M1–M3 in the supplement). Postprocessed values of τ retrieved from the CL51 are broadly in agreement with those from AERONET, and thus can be used to estimate τ in the subcloud layers and at nighttime. If we arbitrarily define a dust outbreak as τ ≥ 0.2, from these data the duration of the 9 March event was approximately 5 h (Fig. 7a, 1835 to 2335 UTC), the 15 March event lasted 5.5 h (Fig. 7b, 2215 to 0345 UTC), and the 15 February event had a duration of 11.17 h (Fig. 7c, 1705 to 0415 UTC), although the latter event was punctuated by distinct periods of τ < 0.2 at 1500, 1930, and 0300 UTC.
Aerosol optical depth τ retrievals and surface wind speeds measured at the field site during the three dust cases. (a)–(c) Time series of τ retrieved from the CL51 (solid line) and the AERONET sun photometer (circles) during each of the three dust outbreaks. (d)–(f) Corresponding measurements of surface wind speeds (blue) and gusts (red-orange), with time periods during which τ ≥ 0.2 indicated by the gray shading.
Citation: Journal of the Atmospheric Sciences 80, 3; 10.1175/JAS-D-22-0128.1
A comparison of measurements of τ and the corresponding surface wind speed Us suggests that, in general, τ ≥ 0.2 when the surface wind speeds and gusts exceed 9 and 17 m s−1, respectively (gray shading in Figs. 7d–f). Although for the 15 February case there are several time periods during which τ > 0.2 but wind speeds are well below 9 m s−1, and when τ < 0.2 but wind speeds are above 9 m s−1 (Fig. 7f). For these cases dust over the field site is emitted from the upwind desert region to the west (see animations M1–M3 in the supplement), and as discussed in section 4, decoupling of τ and Us in the 15 February case may be due to the influence of nonstationary trapped waves.
Measurements of backscatter made from the CL51 ceilometer located at the field site provide information about the vertical structure of the dust storms. Plotted in Fig. 8 are log10 of the ceilometer range corrected backscatter signal within the lower 4 km of the atmosphere for 24-h periods commencing at 1200 UTC on three dates: 9 March (Fig. 8a) and 15 March 2021 (Fig. 8c) and 15 February 2022 (Fig. 8e). Values of log10 backscatter that are greater than 2 are a reasonable indication of the presence of suspended dust based on comparisons with aerosol optical depth retrievals from the collocated sun photometer (Evan et al. 2022a), and backscatter values greater than 2 that are located well above the surface indicated the presence of clouds (e.g., 2–3 km AGL at 2000 UTC 15 March in Fig. 8c).
Ceilometer backscatter and extinction profiles from CL51 measurements made at the field site. Shown are vertical profiles of the log of the CL51 range corrected signal during the dust outbreaks on (a) 9 and 10 Mar and (c) 15 and 16 Mar 2021, and (e) 15 and 16 Feb 2022. The vertical black lines in each panel represent times radiosondes were launched at the site (Fig. 9). (b),(d),(f) Extinction profiles averaged over the time period indicated in the legends (in HH UTC), and corresponding to the days indicated in the adjacent panels.
Citation: Journal of the Atmospheric Sciences 80, 3; 10.1175/JAS-D-22-0128.1
The CL51 data show that for all three cases dust is confined to a layer below 2 km AGL, but that the depth of the dust plume and the vertical distribution of the aerosols vary both between events and within the individual dust outbreaks. For example, the dust outbreak on 9 March is characterized by a plume having depth 1–2 km AGL (Fig. 8a) with extinction values peaking at 500 m AGL (Fig. 8b). For the 15 March case dust is confined to the shallow layer of 300–700 m AGL (Fig. 8c), with extinction peaking at the lowest retrievable level of 100 m AGL (Fig. 8b). Differences in the shapes of the extinction profiles for the 9 and 15 March cases explain why surface PM10 measurements for the 15 March case were far greater than those for 9 March (Figs. 6a,b) although the dust optical depth for these two events are nearly identical in magnitude (Figs. 7a,b). The ceilometer data for 15 February exhibit distinct periods of dust-layer depths, ranging from 600 m to 2 km AGL (Fig. 8e). We consider the factors affecting the vertical distribution of dust in section 4.
c. Terrain-forced flow
Having provided an overview of the synoptic situation for these dust events and examined the physical characteristics of the airborne dust, we next consider the role of orography in generating the high winds that gave rise to the dust outbreaks. The mountain range that lies immediately to the west of the Salton basin (the Peninsular Mountains) is north–south oriented and rises gradually from the Pacific Ocean to peak heights up to 3 km, with steep eastern slopes that plunge into the sub-sea-level Salton basin (Fig. 1b; see also Fig. 5 in Evan 2019). Given the characteristics of the Peninsular Mountains (which hereafter we also refer to as the upwind barrier), wind in the zonal direction is cross barrier and thus westerly flow has the potential to generate strong downslope windstorms in the lee of these mountains (Durran 1990). To elucidate the influence of the orography on the leeside flow we examine radiosondes launched from the field site on each of the days in question and output from WRF simulations of these events.
A profile of potential temperature θ obtained from a radiosonde launched prior to the 9 March dust outbreak at 1500 UTC (0700 local time) shows the remnants of a nocturnal inversion, with θ increasing from 15° to 18°C from the surface to 1.5 km AGL, which is then capped by an approximately 4°C inversion layer, with θ increasing steadily above (Fig. 9a). The corresponding cross-barrier wind speeds u vary between 10 and 20 m s−1 throughout the lower 6 km of the atmosphere (Fig. 9b). A radiosonde released at 2400 UTC on this day (1600 local time), which is during the dust outbreak (Fig. 8a), shows 5°C warming at the surface relative to the 1500 UTC sounding but little change in the 1–1.5-km layer. If we define the top of the convective boundary layer as the height at which θ equals the surface temperature, which is reasonable given that the layer is dry, the depth of the convective boundary layer during the dust outbreak is 1.5 km, which is consistent with the depth of the dust layer during this event (Fig. 8b).
Sounding measurements from radiosondes launched from the field site on (a)–(c) 9 Mar 2021, (d)–(f) 15 Mar 2021, and (g)–(i) 15 Feb 2022. Plotted are radiosonde profiles of (a),(d),(g) potential temperature θ, (b),(e),(h) zonal wind speed u, and (c),(f),(i) balloon ascent rate. Times of the radiosonde launches (HH:MM UTC) are indicated in the legends in (a), (d), and (g).
Citation: Journal of the Atmospheric Sciences 80, 3; 10.1175/JAS-D-22-0128.1
During the 9 March dust outbreak the profile of u can be characterized as consisting of a low-level jet having peak wind speeds of 20 m s−1 from just above the surface to a height of 1 km AGL, and a wind speed minimum of 5 m s−1 at the height of the inversion at 1.5 km AGL (Fig. 9b). The height of the wind speed minimum and 4°C inversion are also located at a minimum in the balloon’s ascent rate, which is in contrast to the more constant ascent rate prior to the dust outbreak (Fig. 9c). Inversions apparent in the profile of θ at heights of 1.5, 3.5, and 4.9 km AGL are coincident with minima in ascent rate and thus the magnitudes of these inversions are affected by the reductions in the radiosonde’s vertical velocity. Minima in the ascent rate also indicate the presence of waves, similar to cases examined in Strauss et al. (2016). The presence of waves is also apparent in measurements from other soundings made during these events (Figs. S6 and S12).
Radiosondes launched immediately prior to and during the dust outbreak on 15 March 2021 show some similar characteristics to those from the 9 March case. The vertical profile of θ prior to the dust outbreak at 2115 UTC (1415 local time, Fig. 8c) suggests a well-mixed boundary layer extending from the surface to approximately 2 km AGL (Fig. 9d) with u near 10 m s−1 throughout this depth (Fig. 9e). In contrast, the sounding made during dust outbreak (2320 UTC, 1620 local time) is accompanied by cooling of approximately 3°C in the lower 500 m of the atmosphere (Fig. 9d) and a pronounced low-level jet characterized by peak wind speeds of 23 m s−1 at heights of 100–300 m AGL and a wind speed minimum of 5 m s−1 at 1.25 km AGL. The radiosonde ascent rates implies wave activity in the atmosphere, with a minimum in ascent rate at 2.8 km AGL (Fig. 9f) that is located at the height of a nearly 10°C inversion (Fig. 9d). We again note that this apparent inversion is heavily influenced by the nearly horizontal motion of the balloon at this height.
The low-level cooling accompanying the onset of high wind speeds helps explain the observed shallow depth of the dust layer on 15 March, relative to the 9 March case (Fig. 8d). These features of the 15 March dust outbreak are similar to those for a dust outbreak that occurred on 22 February 2021, which was generated by spillover precipitation and evaporative cooling over the desert to the west of the research site (Evan et al. 2022c). Here we noted no spillover precipitation for the 15 March case and thus any density-current-like features are due to cold postfrontal downslope flow (Karyampudi et al. 1995; Koch et al. 1991).
For the 15 February 2022 case no radiosondes were launched prior to the dust outbreak, although the radiosondes measurements shown in Figs. 9g–l do correspond to periods of differing heights of the dust plume (Fig. 8e). Profiles of potential temperature made at 2100 UTC (1300 local time) and 2340 UTC (1540 local time) show an inversion just below 2 km AGL (Fig. 9g). For the earlier time we estimate a convective boundary layer depth of 1.6 km, which is consistent with the depth of the dust layer averaged from 2000 to 2300 UTC (Fig. 8f). For the 2340 UTC sounding there is relative cooling of approximately 1.5°C in the lower 1.8 km of the atmosphere, and 2.0°C at the surface. This change in the θ profile suggests that the depth of the convective boundary layer is reduced to 1 km AGL, consistent with the depth of the dust layer averaged from 2300 to 0300 UTC (Fig. 8f).
Similar to the 9 and 15 March cases, zonal wind speeds from radiosondes launched during the dust outbreak on 15 February show low-level jets, with speed maxima of 18 m s−1 located at heights of 400–500 m AGL, and wind speed minima of 8 m s−1 at 1.5–1.75 km AGL (Fig. 9j). For these cases we also find minima in radiosonde ascent rates that are coincident with inversions present in the θ profiles, including at 1.9, 3.9, and 5.4 km AGL for the 2100 UTC sounding (1300 local time), and 1.7, 4.5, and 5.3 km AGL at 2340 UTC (1540 local time, Fig. 9k), again reflecting the presence of waves in the overlying atmosphere.
We again utilize radiosondes made from the NKX sounding station located near the coast (Figs. 1 and 3) in order to understand the factors that give rise to these downslope windstorms via examination of radiosondes released at 2400 UTC from this location and the nearest in time radiosondes released from the field site near the Salton Sea (Fig. 10). According to Mayr and Armi (2010) leeside flow will plunge to the floor of the basin if the potential temperature of the air flowing over the ridge is cooler than that of the down-barrier surface. The heights of the ridgeline upwind of the field site are in the range of 1.25–1.75 km MSL (gray shaded band in Fig. 10), and the upwind (NKX) potential temperatures at these heights (Figs. 10a–c, light blue) are all lower than the downwind values of θ below 1.25 km (Figs. 10a–c, rust). Vertical profiles of wind speed from the NXK soundings suggest upwind orographic flow blocking, as evidenced by wind speeds below 1.25 km in the range of 5–12 m s−1 that are in contrast to the high wind speeds downwind of the barrier (Figs. 10d–f). Above the heights of the ridge the upwind and downwind wind speed profiles are similar, with the exception of the jet at 1.75–2.5 km in the 15 February downwind profile (Fig. 10f), which is due to the influence of wave activity on the radiosonde ascent rate. These differences in the soundings upwind and downwind of the barrier are consistent with isentropic drawdown of the cross-barrier flow at or above the height of the ridgeline and a leeside downslope windstorm (Durran 1990).
Radiosonde profiles of (top) θ and (bottom) wind speed from San Diego, California (NKX), which is upwind of the barrier (light blue) and from the field site that is located near the Salton Sea (rust). The heights of the mountain ridge represented by the gray shaded band in each panel. Profiles are shown for the 2400 UTC soundings from San Diego and the radiosondes launched closest to this time near the Salton Sea (i.e., the later sounding times in Fig. 9) for (a),(d) 9 March and (b),(e) 15 Mar 2021, and (c),(f) 15 Feb 2022.
Citation: Journal of the Atmospheric Sciences 80, 3; 10.1175/JAS-D-22-0128.1
d. Numerical simulations with WRF
To provide broader context to the in situ measurements we also examine output from numerical simulations using WRF, focusing on model output along the 33.25°N latitude transect for the innermost model domain (Fig. 2). The WRF simulations reproduced several aspects of the surface and upper air measurements, including strong westerly surface wind speeds during the dust events (Fig. S1). However, at least at the field site, the simulated timing of the onset and termination of high wind speeds did not line up with observations, and for several cases waves in the model appeared to be out of phase with wave activity implied by changes in the radiosonde ascent rate (Figs. S2–S13). As such, WRF output is used to understand the general behavior of the downslope flow and trapped waves in the Salton basin, rather than to explain the timing of specific aspects of these events.
Hovmöller diagrams of vertical velocity w averaged over the 3–4-km layer during 24-h time periods starting at 1200 UTC on 9 March (Fig. 11a) and 15 March 2021 (Fig. 11c) and 15 February 2022 (Fig. 11e) indicate the presence of trapped lee waves during the periods of observed high winds and dust (Fig. 7). For all three cases and during the entire 24-h time period downslope flow is simulated along the leeside slopes of the upwind barrier (the barrier ridge is located at the 0-km point on the horizontal axis and flow downwind of the barrier is located at positive horizontal distances, a transect of the orography is found in Fig. 12), and then vertical ascent at 10 km distance from the barrier. This type of plunging flow and downwind jump has been the focus of research on high wind events and dust storms in the Owens valley (e.g., Grubišić et al. 2008). Indeed, similarly constructed Hovmöller diagrams of u at the lowest model level indicate the strongest surface winds (u > 20 m s−1) along the leeside slopes for all three cases (Figs. 11b,d,f). However, distinct from the narrow Owens valley, in the Salton basin the terrain of the first 35 km downwind of the barrier is vegetated and generally nonemissive, and as such the high winds associated with the flow at the base of the barrier do not produce dust here.
Hovmöller diagrams of (a),(c),(e) 3–4-km-height averaged vertical velocity w and (b),(d),(f) surface zonal wind speed us along the 33.25°N latitude transect during the (a),(b) 9 Mar 2021, (c),(d) 15 Mar 2021, and (e),(f) 15 Feb 2022 dust outbreaks. The upward-pointing arrows in (a) and (b) indicate the location of the field site. Reference orography along this transect can be found in Fig. 12.
Citation: Journal of the Atmospheric Sciences 80, 3; 10.1175/JAS-D-22-0128.1
WRF output along the 33.25°N latitude transect averaged over simulation times (a)–(c) 2100–2300 UTC 9 Mar 2021, (d)–(f) 2200–2400 UTC 15 Mar 2021, and (g)–(i) 2200–2400 UTC 15 Feb 2022. Horizontal distance is given in km from the peak of the upwind orography. Plotted in (a), (d), and (g) are vertical wind speeds w at contour intervals of 1 m s−1, with warm colors representing positive w and cool colors representing negative w, and where the 0 m s−1 isotach is not plotted. Plotted in (b), (e), and (h) are lines of constant θ at 1°C intervals. Shown in (c), (f), and (i) is u, where the white shaded region at the downwind base of the orography indicates reversed flow (u < 0). The downward-pointing arrow in (a) represents the location of the field site. The red horizontal line on the bottom axis in all panels represents the approximate locations of dust emission that are upwind of the field site.
Citation: Journal of the Atmospheric Sciences 80, 3; 10.1175/JAS-D-22-0128.1
A distance–height transect of model output vertical velocity w (Figs. 12a,d,g) and dry isentropes (Figs. 12b,e,h), averaged over 2-h time periods during which the waves are approximately stationary, indicate the existence of trapped waves in all three cases. The weakest wave activity is seen in the WRF output for 9 March, where the magnitude of the vertical wind speeds drop below 1 m s−1 at a distance of approximately 60 km from the mountain ridge (Fig. 12a). The 15 March case exhibits the strongest wave activity, with waves of quasi-regular wavelength 20 km and vertical velocity magnitudes as large as 5 m s−1 at a barrier distance of 85 km (Fig. 12d). The 15 February case also shows strong wave activity throughout the model domain, but of smaller magnitude and longer wavelength than that for 15 March (Figs. 12g). For all three cases the waves are evanescent above approximately 6-km height (Figs. 12b,e,h), due to changes in static stability and vertical wind shear in the flow upstream of the orography (Fig. 10).
Relevant to understanding the influence of trapped waves on dust emission and transport is their effect on surface wind speed us. First, from the Hovmöller diagrams in Fig. 11 the strongest wind speeds are in general found along the leeside slopes, with weak and even reversed flow just downwind of the barrier base, indicative of flow separation and a rotor circulation (Doyle and Durran 2002). Farther downwind of the barrier the strongest surface wind speeds (us > 20 m s−1) are associated with the presence of trapped waves. For example, in the 15 February case plunging flow along the leeside slopes (0–10 km) produce horizontal surface wind speeds near 30 m s−1 from 1200 to 0000 UTC (Fig. 11f). Prior to the development of trapped waves at approximately 1800 UTC (Fig. 11e) surface wind speeds at barrier distances greater than 20 km are below 10 m s−1. As lee waves develop surface wind speeds greater than 25 m s−1 are found as far as 90 km from the barrier. In general and for these cases, since dust emission primarily occurs at barrier distances greater than 35 km, significant dust uplift in the basin would only occur after lee wave onset.
The effect of wave activity on surface wind speed us is apparent in the cross sections of zonal wind speed (Figs. 12c,f,i). Perturbations in us are out of phase with horizontal gradients in w and are in phase with θ, which is due to surface pressure minima under the regions of strongest upward vertical velocity, and surface pressure maxima located under the strongest downdrafts (Nappo 2013). For all three cases the strongest surface wind speeds are all found under the wave troughs. Although the speed of the plunging flow along the leeside slopes is very similar for all three cases, surface wind speeds farther downwind of the barrier (distances greater than 20 km) are the weakest in the 9 March case, in which the trapped waves are less pronounced and dissipate at barrier distances greater than 50 km, and are the strongest downwind during the 15 March case, in which the waves are still coherent at barrier distances greater than 80 km.
4. Discussion
Plots of l2 for the 2400 UTC soundings made upwind of the barrier at the NKX site (see location in Fig. 1) show a reduction in l2 with height for all three cases (Fig. 13). For 9 March, l2 peaks low in the atmosphere, at a height of 1.3 km (Fig. 13a) and reaches a minimum at 4 km. For the other two cases, l2 peaks slightly higher in the atmosphere at 2 km, and reaches a minimum at 5.8 km (Fig. 13c) and 4.5 km (Fig. 13e). Profiles of the first term on the right-hand side of Eq. (1) suggest that variations in l2 are primarily driven by reductions in static stability above 2 km (Figs. 13b,d,f), since in all three cases
Changes in the Scorer parameter with height. (left) The Scorer parameter l2 [Eq. (1)] calculated from the 2400 UTC soundings made from the NKX station (blue) and output from the WRF simulations (rust) for the dates indicated at left. (right) The l2 stability term [first term on the right-hand side of Eq. (1)] for the same data. Only values above 1 km are shown due to the blocked flow below this height (e.g., Fig. 10).
Citation: Journal of the Atmospheric Sciences 80, 3; 10.1175/JAS-D-22-0128.1
In the model the reductions in l2 with height are not as large as in the observations (Figs. 13a,c,e). This difference is due to discrepancies between the observed and simulated shear profiles as there is generally good agreement in the l2 stability terms. We note that for the 9 March case the modeled stability term does not drop off as strongly with height as does that from observations (Fig. 13b), and thus it is plausible that the model is under predicting trapped wave activity for the three cases considered here.
Based on the measurements and model output presented here we suggest several characteristics of dust storms generated by trapped waves. The first is the presence of a low-level jet (Figs. 10d–f), which is relevant in terms of dust production, advection, and dispersion due to the strong wind speeds characterizing the jet, and the potential for enhanced turbulence production in the shear zones above and below the jet nose. Pressure perturbations associated with wave phase generate positive and negative horizontal wind speed perturbations under the wave troughs and crests, respectively (Durran 1986), resulting in vertical profiles of horizontal wind speeds u that resemble a low-level jet at the base of the wave trough or at the surface under a wave crest. The effect of wave phase on vertical profiles of wind speed can be readily seen in the output from the WRF simulations for the 15 March case (Figs. 12d–f), averaged from 2200 to 2400 UTC. Focusing on one cycle of the simulated wave over barrier distances of 20–40 km (Fig. 14a), the simulated zonal (i.e., cross-barrier) wind speed u is stronger under the wave crest than under the trough, from the surface up to a height of 2 km AGL (Fig. 14b), which is the height where the isentropes above the trough start to spread vertically. Under the wave trough u increases by 10 m s−1 from the surface to the base of the wave at 1 km AGL. Under the crest there is a local maximum in u at approximately 500 m AGL above which u decreases by 2 m s−1 to the local minimum at 1.3 km AGL. As such, the low-level jet under the wave trough is more pronounced than that for the crest and has a nose located at the base of the wave, while that under the crest is weaker with a nose located close to the surface.
Changes in model zonal wind speed u and potential temperature θ with height for different wave phases and proximity to the surface. (a),(d) Contours of isentropic surfaces at 1°C intervals from the WRF output shown in Fig. 12e (15 Mar 2021 case) at barrier distances of (a) 20–40 and (d) 60–83 km, where the vertical blue and rust colored lines indicate the locations of the wave troughs and ridges used to generate the profiles in the other panels. (b),(e) Profiles of cross-barrier wind speed u corresponding to the wave troughs (blue) and ridges (rust). (c),(f) As in (b) and (e), but for θ, and where the horizontal lines indicate the model-calculated height of the convective boundary layer (CBL).
Citation: Journal of the Atmospheric Sciences 80, 3; 10.1175/JAS-D-22-0128.1
For the same WRF simulation but at barrier distances of 60–83 km (Fig. 14d), possibly more representative of the environment over the field site, the simulated wave is evanescent to the surface. Here u is greater under the wave trough than the crest up to a height of 4 km AGL (Fig. 14e), which for the trough is the height above which the isentropic surfaces start to spread vertically. While there is no obviously discernible low-level jet under the wave trough, under the crest there is a greater than 10 m s−1 reduction in u from the local maximum at 500 m up to the minimum at 2.5 km AGL, above which the isentropes become more tightly packed, signifying the wave base. The similarity between the low-level jet under the wave crest in Fig. 14e and the wind speed profiles in the soundings made during the dust outbreaks (Figs. 9b,e,h) raise the possibility that the site is often located under wave crests during trapped wave events.
We suggest that another characteristic of dust storms generated by trapped lee waves is the variable convective boundary layer and thus dust-layer depths, which are dependent upon the phase and proximity to the surface of the overlying wave. Returning to the WRF output from the 15 March case and barrier distances of 20–40 km (Fig. 14a), under the wave trough isentropes are displaced downward toward the surface, resulting in a modeled boundary layer height of 1 km AGL (Fig. 14c). In contrast, under the wave crest isentropic surfaces are displaced upward such that θ is little changed from the surface up to nearly 2 km AGL, with a corresponding boundary layer height of 2.4 km AGL. When considering distances of 60–83 km from the barrier (Fig. 14d) there is little difference in the vertical distribution of θ under the wave trough and crest in the lower 1 km of the atmosphere due to the proximity of the wave to the surface, with each exhibiting similar boundary layer heights of 0.7 and 0.8 km AGL, respectively (Fig. 14f), which are shallower than those for the previous case. The relatively shallow simulated boundary layers in Fig. 14f may explain why for the three cases considered here the observed dust-layer depths are shallow (Fig. 8), and the surface PM10 concentrations are high (Fig. 6).
We further consider the effect of wave phase on dust-layer depth via simulations with WRF-Chem. A transect of dust concentration from WRF-Chem for the 15 March case at 2330 UTC and for the first 70 km downwind of the barrier indicates that the depth of the dust layer closely follows the curvature of the isentropic surfaces that define trapped wave base (Fig. 15a). Furthermore, the highest dust concentrations are found under the wave crests, where the simulated zonal and cross-barrier wind speeds are near zero or negative (Fig. 12f) and the boundary layer turbulent kinetic energy is large (not shown), implying that the areas under the wave crests are regions of strong vertical diffusion and weak down-barrier transport of dust, explaining why the isopleths of high dust concentrations (e.g., >0.2 mg m−3) increase with barrier distance.
Dust simulated by WRF-Chem at 2330 UTC 15 Mar 2021. (a) Isentropic surfaces at 1°C intervals (contours) and dust concentration (mg m−3) along the 33.25°N zonal transect, in km downwind of the barrier crest. (b) The horizontal distribution of dust mass path (g m−2) along the transect in (a), but in zonal units of °E. Black contour lines indicate topography intervals of 250 m, the thick black line indicates the Salton Sea shoreline, and the white horizontal line is the latitude of the transect in (a). (c) As in (b), but for the average dust emission over the preceding 30 min (g m−2 s−1).
Citation: Journal of the Atmospheric Sciences 80, 3; 10.1175/JAS-D-22-0128.1
A map of the horizontal structure of dust mass path, which is the vertically integrated concentration, also for 2330 UTC on this date (Fig. 15b) shows coherent northwest–southeast oriented wave fronts of high and low dust mass path that closely follow the orientation of the upwind topography. As such, in addition to depth of the dust layer, trapped waves have a strong effect on the horizontal distribution of dust concentration. A map of the corresponding surface dust emission flux (Fig. 15c) does not clearly show any resemblance to the structure of the waves, owing to the dominant influence of surface characteristics on dust emission, suggesting that the spatial structures of dust concentration and mass path are largely the result of advection and diffusion rather than the spatial pattern of emission. We also note that under the wave crests the cross-barrier wind speeds are weak but the along-barrier wind speeds are northerly (not shown), raising the possibility of meridional dust advection there.
We again note that while these WRF-Chem simulations are useful in terms of elucidating the general characteristics of dust storms generated by trapped waves, they are of limited use in terms of understanding the specific distribution of dust in the region during these events. Eyewitness accounts and GOES-R and Roundshot camera animations (M1–M3 in the supplement) show that during these events dust is mainly emitted from the low-lying desert regions (i.e., barrier distances greater than 40 km in Fig. 15a) whereas the model shows little to no emission in this area (Fig. 15c, 33.2° to 33.3°N, −116.2° to −116°E). Ongoing work suggests that apparent unrealistic distribution of dust emission is at least in part due to erroneous land-surface-type classification.
Last, our results imply that a third characteristic of dust storms generated by trapped waves is that wave-forced wind speed perturbations, and thus dust emission, can occur far downwind of the barrier. Output from the WRF simulations indicate that surface wind speed perturbations associated with trapped waves occur as far as 100 km downwind of the barrier (Figs. 11, 12). Radiosondes also indicate the presence of waves downwind of the field site during all three cases (Fig. 9), where plots of balloon height and ascent rate as a function of zonal distance from the field site imply that waves are found at barrier distances greater than 100 km (e.g., Figs S6, S8, and S13).
5. Conclusions
Observations of three dust outbreaks that occurred in the northwestern Sonoran Desert indicated that these storms were all associated with the presence of trapped lee waves generated by a north–south oriented mountain range. Reanalysis demonstrated that for each case cross-barrier flow was directed over the region by way of a synoptic-scale low pressure trough transitioning through the area (Fig. 3). Surface meteorological measurements showed that during trough passage flow over a field site located near the western shoreline of the Salton Sea (Fig. 1) was westerly with wind speeds and gusts exceeding 10 and 20 m s−1, respectively (Fig. 5). Measurements of PM10 (Fig. 6) and animations from a Roundshot camera and GOES-17 (supplemental materials M1–M3) indicated the presence of dust across the region, and aerosol optical depth retrievals from a sun photometer and a ceilometer exhibited values greater than 0.3 during the dust outbreaks. Backscatter profiles from the ceilometer suggested that the depths of the dust layers ranged from 700 m to 2 km (Fig. 8). Radiosondes released prior to and during the dust events suggested that the high winds were associated with a shallow convective boundary layer, one factor in generating the shallow dust layers, and the presence of a jet in the lower 1.5 km of the atmosphere (Fig. 9). Radiosonde ascent rates implied the presence of trapped waves in the environment downwind of the field site (Fig. 9, Figs. S6, S8, and S10–13), consistent with numerical simulations conducted with the WRF Model showing that each of the dust-producing high wind events were at some point associated with the presence of trapped lee waves (Fig. 11), resulting in positive surface wind speed perturbations far downwind of the wave-generating barrier (Fig. 12).
We highlighted several meteorological aspects of the observed and simulated trapped waves that are relevant to understanding the characteristics of the concurrent dust outbreaks. These include the presence of a low-level jet whose depth and speed is affected by wave phase and vertical structure, dust-layer depths and concentrations that are also dependent upon these factors, and high wind speeds and dust emission more than 100 km downwind the wave source. Output from WRF-Chem provided corroborating evidence that the depth of the dust layer is strongly tied to wave phase, with the model showing the highest dust concentrations under the wave crests. Direct observational evidence to evaluate many aspects of the wave-forced dust storm characteristics (e.g., the relationship between wave phase and depth of the dust layer) would require measurements of aerosols and meteorology at different wave phases and at concurrent times, something that is not currently possible given the available instrumentation at this single field site.
Inversions upwind and near the heights of the ridge of the Peninsular Mountains were noted for the 15 March 2021 and 15 February 2022 cases (Fig. 10), as well as in the cases examined in Evan et al. (2022c) and Evan (2019). As such, trapped waves are likely a common feature of strong cross barrier flow and dust outbreaks in the region. Observations and modeling from the Owens valley suggest, however, that the Salton Sea is not unique in this regard (e.g., Grubišić and Billings 2007). More work to evaluate the role of trapped waves on dust emission in other dust-emitting regions is warranted, especially since climate models do not directly simulate nor parameterize trapped waves.
The Salton Sea is rapidly drying, and thus the area of exposed playa and potential for increasing dust emission is growing. The Salton Sea sits immediately downwind of the field site, and as such trapped waves have the ability to generate high wind speeds and dust over the growing playa surfaces. It is not clear how the drying of the sea and the resultant changes in the surface temperature and sensible and latent heat fluxes will feed back onto wave activity. It is possible that a warming surface will heat the overlying atmosphere resulting in a reduction of wave amplitude (Jiang et al. 2006), although this effect could also increase the strength of the surface winds by allowing isentropes at the barrier level to more frequently reach the downwind surface. It is also unknown how drying of the sea may affect the depth of the dust layer; while increased surface heating implies a deeper convective boundary layer, the interaction of surface warming with wave activity may increase the near-surface stability. It is also plausible that radiative heating by the dust will in turn feed back onto the wave characteristics. Given the rapid environmental change occurring in this region and the health impacts of exposure to dust on the community (Frie et al. 2017, 2019; Jones and Fleck 2020; Biddle et al. 2022), more work to elucidate the impacts of the drying Salton Sea on the region’s meteorology and air quality is warranted.
Acknowledgments.
Funding for this work was provided by NSF Award AGS-1833173. We thank Trinity Robinson and Tyler Barbero for their assistance on this project, and three anonymous reviewers for their comments on a previous version of this manuscript.
Data availability statement.
Salton Sea AERONET data are available via aeronet.gsfc.nasa.gov/, surface synoptic station data are at mesowest.utah.edu/, GOES-17 satellite data are at www.avl.class.noaa.gov, NEXRAD data are at mesonet.agron.iastate.edu, surface PM10 measurements are from www.arb.ca.gov/aqmis2/aqdselect.php, NARR output is available from psl.noaa.gov, and GFS analysis is available from www.nco.ncep.noaa.gov/pmb/products/gfs/. The soundings, surface meteorological data, and CL51 backscatter and extinction profiles used in this manuscript are permanently archived at https://doi.org/10.6075/J0XP7547 (Evan et al. 2022b).
REFERENCES
Biddle, T. A., and Coauthors, 2021: Salton Sea aerosol exposure in mice induces a pulmonary response distinct from allergic inflammation. Sci. Total Environ., 792, 148450, https://doi.org/10.1016/j.scitotenv.2021.148450.
Biddle, T. A., R. Chakraborty, Q. Li, M. Maltz, J. Gerrard, and D. D. Lo, 2022: The drying Salton Sea and asthma: A perspective on a “natural” disaster. Calif. Agric., 76, 27–36, https://doi.org/10.3733/ca.2022a0003.
Buck, B. J., J. King, and V. Etyemezian, 2011: Effects of salt mineralogy on dust emissions, Salton Sea, California. Soil Sci. Soc. Amer. J., 75, 1971–1985, https://doi.org/10.2136/sssaj2011.0049.
Burr, A. C., and Coauthors, 2021: Lung inflammatory response to environmental dust exposure in mice suggests a link to regional respiratory disease risk. J. Inflammation Res., 14, 4035–4052, https://doi.org/10.2147/JIR.S320096.
Chen, F., and J. Dudhia, 2001: Coupling an advanced land surface–hydrology model with the Penn State–NCAR MM5 modeling system. Part I: Model implementation and sensitivity. Mon. Wea. Rev., 129, 569–585, https://doi.org/10.1175/1520-0493(2001)129<0569:CAALSH>2.0.CO;2.
Chen, S.-H., and W.-Y. Sun, 2002: A one-dimensional time dependent cloud model. J. Meteor. Soc. Japan, 80, 99–118, https://doi.org/10.2151/jmsj.80.99.
Chin, M., R. B. Rood, S.-J. Lin, J.-F. Müller, and A. M. Thompson, 2000: Atmospheric sulfur cycle simulated in the global model GOCART: Model description and global properties. J. Geophys. Res., 105, 24 671–24 687, https://doi.org/10.1029/2000JD900384.
Choobari, O. A., P. Zawar-Reza, and A. Sturman, 2014: The global distribution of mineral dust and its impacts on the climate system: A review. Atmos. Res., 138, 152–165, https://doi.org/10.1016/j.atmosres.2013.11.007.
De Wekker, S. F., and S. D. Mayor, 2009: Observations of atmospheric structure and dynamics in the Owens valley of California with a ground-based, eye-safe, scanning aerosol lidar. J. Appl. Meteor. Climatol., 48, 1483–1499, https://doi.org/10.1175/2009JAMC2034.1.
Doyle, J. D., and D. R. Durran, 2002: The dynamics of mountain-wave-induced rotors. J. Atmos. Sci., 59, 186–201, https://doi.org/10.1175/1520-0469(2002)059<0186:TDOMWI>2.0.CO;2.
Durran, D. R., 1986: Another look at downslope windstorms. Part I: The development of analogs to supercritical flow in an infinitely deep, continuously stratified fluid. J. Atmos. Sci., 43, 2527–2543, https://doi.org/10.1175/1520-0469(1986)043<2527:ALADWP>2.0.CO;2.
Durran, D. R., 1990: Mountain waves and downslope winds. Atmospheric Processes over Complex Terrain, Springer, 59–81, https://doi.org/10.1007/978-1-935704-25-6_4.
Durran, D. R., 2003: Lee waves and mountain waves. Encyclopedia of Atmospheric Sciences, J. R. Holton, J. Pyle, and J. A. Curry, Eds., Elsevier Science, 1161–1169.
Evan, A. T., 2019: Downslope winds and dust storms in the Salton basin. Mon. Wea. Rev., 147, 2387–2402, https://doi.org/10.1175/MWR-D-18-0357.1.
Evan, A. T., C. Flamant, M. Gaetani, and F. Guichard, 2016: The past, present and future of African dust. Nature, 531, 493–495, https://doi.org/10.1038/nature17149.
Evan, A. T., B. Walkowiak, and R. Frouin, 2022a: On the misclassification of dust as cloud at an AERONET site in the Sonoran desert. J. Atmos. Oceanic Technol., 39, 181–191, https://doi.org/10.1175/JTECH-D-21-0114.1.
Evan, A. T., R. Frouin, A. Kuwano, T. Barbero, S. Wynn, and T. Robinson, 2022b: Meteorological and aerosol measurements near the Salton Sea, California. UC San Diego Library Digital Collections, accessed 31 July 2022, https://doi.org/10.6075/J0XP7547.
Evan, A. T., W. Porter, R. Clemesha, A. Kuwano, and R. Frouin, 2022c: Measurements of a dusty density current in the western Sonoran desert. J. Geophys. Res. Atmos. 127, e2021JD035830, https://doi.org/10.1029/2021JD035830.
Fast, J. D., W. I. Gustafson Jr., R. C. Easter, R. A. Zaveri, J. C. Barnard, E. G. Chapman, G. A. Grell, and S. E. Peckham, 2006: Evolution of ozone, particulates, and aerosol direct radiative forcing in the vicinity of Houston using a fully coupled meteorology-chemistry-aerosol model. J. Geophys. Res., 111, D21305, https://doi.org/10.1029/2005JD006721.
Fernald, F. G., 1984: Analysis of atmospheric lidar observations: Some comments. Appl. Opt., 23, 652–653, https://doi.org/10.1364/AO.23.000652.
Field, J. P., and Coauthors, 2010: The ecology of dust. Front. Ecol. Environ., 8, 423–430, https://doi.org/10.1890/090050.
Frie, A. L., J. H. Dingle, S. C. Ying, and R. Bahreini, 2017: The effect of a receding saline lake (the Salton Sea) on airborne particulate matter composition. Environ. Sci. Technol., 51, 8283–8292, https://doi.org/10.1021/acs.est.7b01773.
Frie, A. L., and Coauthors, 2019: Dust sources in the Salton Sea basin: A clear case of an anthropogenically impacted dust budget. Environ. Sci. Technol., 53, 9378–9388, https://doi.org/10.1021/acs.est.9b02137.
Giles, D. M., and Coauthors, 2019: Advancements in the Aerosol Robotic Network (AERONET) version 3 database—Automated near-real-time quality control algorithm with improved cloud screening for sun photometer aerosol optical depth (AOD) measurements. Atmos. Meas. Tech., 12, 169–209, https://doi.org/10.5194/amt-12-169-2019.
Giles, J., 2005: The dustiest place on Earth. Nature, 434, 816–819, https://doi.org/10.1038/434816a.
Ginoux, P., M. Chin, I. Tegen, J. M. Prospero, B. Holben, O. Dubovik, and S.-J. Lin, 2001: Sources and distributions of dust aerosols simulated with the GOCART model. J. Geophys. Res., 106, 20 255–20 273, https://doi.org/10.1029/2000JD000053.
Ginoux, P., J. M. Prospero, T. E. Gill, N. C. Hsu, and M. Zhao, 2012: Global-scale attribution of anthropogenic and natural dust sources and their emission rates based on MODIS deep blue aerosol products. Rev. Geophys., 50, RG3005, https://doi.org/10.1029/2012RG000388.
Gläser, G., P. Knippertz, and B. Heinold, 2012: Orographic effects and evaporative cooling along a subtropical cold front: The case of the spectacular Saharan dust outbreak of March 2004. Mon. Wea. Rev., 140, 2520–2533, https://doi.org/10.1175/MWR-D-11-00315.1.
Grell, G. A., 1993: Prognostic evaluation of assumptions used by cumulus parameterizations. Mon. Wea. Rev., 121, 764–787, https://doi.org/10.1175/1520-0493(1993)121<0764:PEOAUB>2.0.CO;2.
Grell, G. A., and D. Dévényi, 2002: A generalized approach to parameterizing convection combining ensemble and data assimilation techniques. Geophys. Res. Lett., 29, 1693, https://doi.org/10.1029/2002GL015311.
Grell, G. A., S. E. Peckham, R. Schmitz, S. A. McKeen, G. Frost, W. C. Skamarock, and B. Eder, 2005: Fully coupled “online” chemistry within the WRF Model. Atmos. Environ., 39, 6957–6975, https://doi.org/10.1016/j.atmosenv.2005.04.027.
Grubišić, V., and B. J. Billings, 2007: The intense lee-wave rotor event of Sierra Rotors IOP 8. J. Atmos. Sci., 64, 4178–4201, https://doi.org/10.1175/2006JAS2008.1.
Grubišić, V., and Coauthors, 2008: The Terrain-Induced Rotor Experiment: A field campaign overview including observational highlights. Bull. Amer. Meteor. Soc., 89, 1513–1534, https://doi.org/10.1175/2008BAMS2487.1.
Holben, B. N., and Coauthors, 1998: AERONET—A federated instrument network and data archive for aerosol characterization. Remote Sens. Environ., 66, 1–16, https://doi.org/10.1016/S0034-4257(98)00031-5.
Horel, J., and Coauthors, 2002: MesoWest: Cooperative mesonets in the western United States. Bull. Amer. Meteor. Soc., 83, 211–225, https://doi.org/10.1175/1520-0477(2002)083<0211:MCMITW>2.3.CO;2.
Huneeus, N., and Coauthors, 2011: Global dust model intercomparison in AeroCom phase I. Atmos. Chem. Phys., 11, 7781–7816, https://doi.org/10.5194/acp-11-7781-2011.
Iacono, M. J., J. S. Delamere, E. J. Mlawer, M. W. Shephard, S. A. Clough, and W. D. Collins, 2008: Radiative forcing by long-lived greenhouse gases: Calculations with the AER radiative transfer models. J. Geophys. Res., 113, D13103, https://doi.org/10.1029/2008JD009944.
IID, 2016: Salton Sea air quality mitigation program. SSAQT Tech. Rep., 281 pp.
Janjić, Z. I., 1994: The step-mountain eta coordinate model: Further developments of the convection, viscous sublayer, and turbulence closure schemes. Mon. Wea. Rev., 122, 927–945, https://doi.org/10.1175/1520-0493(1994)122<0927:TSMECM>2.0.CO;2.
Janjić, Z. I., 2001: Nonsingular implementation of the Mellor-Yamada level 2.5 scheme in the NCEP Meso model. NCEP Office Note 437, 61 pp.
Jiang, H., J. T. Farrar, R. C. Beardsley, R. Chen, and C. Chen, 2009: Zonal surface wind jets across the Red Sea due to mountain gap forcing along both sides of the Red Sea. Geophys. Res. Lett., 36, L19605, https://doi.org/10.1029/2009GL040008.
Jiang, Q., J. D. Doyle, and R. B. Smith, 2006: Interaction between trapped waves and boundary layers. J. Atmos. Sci., 63, 617–633, https://doi.org/10.1175/JAS3640.1.
Jiang, Q., M. Liu, and J. D. Doyle, 2011: Influence of mesoscale dynamics and turbulence on fine dust transport in Owens valley. J. Appl. Meteor. Climatol., 50, 20–38, https://doi.org/10.1175/2010JAMC2522.1.
Jin, Y., and Coauthors, 2015: Ceilometer calibration for retrieval of aerosol optical properties. J. Quant. Spectrosc. Radiat. Transfer, 153, 49–56, https://doi.org/10.1016/j.jqsrt.2014.10.009.
Jones, B. A., and J. Fleck, 2020: Shrinking lakes, air pollution, and human health: Evidence from California’s Salton Sea. Sci. Total Environ., 712, 136490, https://doi.org/10.1016/j.scitotenv.2019.136490.
Karyampudi, V. M., S. E. Koch, C. Chen, J. W. Rottman, and M. L. Kaplan, 1995: The influence of the Rocky Mountains on the 13–14 April 1986 severe weather outbreak. Part II: Evolution of a prefrontal bore and its role in triggering a squall line. Mon. Wea. Rev., 123, 1423–1446, https://doi.org/10.1175/1520-0493(1995)123<1423:TIOTRM>2.0.CO;2.
Kim, K.-M., and Coauthors, 2021: Modeling Asian dust storms using WRF-Chem during the DRAGON-Asia field campaign in April 2012. J. Geophys. Res. Atmos., 126, e2021JD034793, https://doi.org/10.1029/2021JD034793.
Knippertz, P., 2014: Meteorological aspects of dust storms. Mineral Dust, P. Knippertz and J. Stuut, Eds., Springer, 121–147, https://doi.org/10.1007/978-94-017-8978-3_6.
Knippertz, P., C. Deutscher, K. Kandler, T. Müller, O. Schulz, and L. Schütz, 2007: Dust mobilization due to density currents in the Atlas region: Observations from the Saharan Mineral Dust Experiment 2006 field campaign. J. Geophys. Res., 112, D21109, https://doi.org/10.1029/2007JD008774.
Koch, S. E., P. B. Dorian, R. Ferrare, S. H. Melfi, W. C. Skillman, and D. Whiteman, 1991: Structure of an internal bore and dissipating gravity current as revealed by Raman lidar. Mon. Wea. Rev., 119, 857–887, https://doi.org/10.1175/1520-0493(1991)119<0857:SOAIBA>2.0.CO;2.
Kok, J. F., and Coauthors, 2014: An improved dust emission model—Part 1: Model description and comparison against measurements. Atmos. Chem. Phys., 14, 13 023–13 041, https://doi.org/10.5194/acp-14-13023-2014.
Kok, J. F., D. S. Ward, N. M. Mahowald, and A. T. Evan, 2018: Global and regional importance of the direct dust-climate feedback. Nat. Commun., 9, 241, https://doi.org/10.1038/s41467-017-02620-y.
LeGrand, S. L., C. Polashenski, T. W. Letcher, G. A. Creighton, S. E. Peckham, and J. D. Cetola, 2019: The AFWA dust emission scheme for the GOCART aerosol model in WRF-Chem v3.8.1. Geosci. Model Dev., 12, 131–166, https://doi.org/10.5194/gmd-12-131-2019.
Marcos, C. R., J. L. Gómez-Amo, C. Peris, R. Pedrós, M. P. Utrillas, and J. A. Martínez-Lozano, 2018: Analysis of four years of ceilometer-derived aerosol backscatter profiles in a coastal site of the western Mediterranean. Atmos. Res., 213, 331–345, https://doi.org/10.1016/j.atmosres.2018.06.016.
Marsham, J. H., P. Knippertz, N. S. Dixon, D. J. Parker, and G. M. Lister, 2011: The importance of the representation of deep convection for modeled dust-generating winds over West Africa during summer. Geophys. Res. Lett., 38, L16803, https://doi.org/10.1029/2011GL048368.
Marticorena, B., and G. Bergametti, 1995: Modeling the atmospheric dust cycle: 1. Design of a soil-derived dust emission scheme. J. Geophys. Res., 100, 16 415–16 430, https://doi.org/10.1029/95JD00690.
Mayr, G. J., and L. Armi, 2010: The influence of downstream diurnal heating on the descent of flow across the Sierras. J. Appl. Meteor. Climatol., 49, 1906–1912, https://doi.org/10.1175/2010JAMC2516.1.
Mesinger, F., and Coauthors, 2006: North American Regional Reanalysis. Bull. Amer. Meteor. Soc., 87, 343–360, https://doi.org/10.1175/BAMS-87-3-343.
Miller, P., M. Williams, and T. Mote, 2021: Modeled atmospheric optical and thermodynamic responses to an exceptional trans-Atlantic dust outbreak. J. Geophys. Res., 126, e2020JD032909, https://doi.org/10.1029/2020JD032909.
Monin, A. S., and A. M. Obukhov, 1954: Basic laws of turbulent mixing in the atmosphere near the ground. Tr. Geofiz. Inst., Akad. Nauk SSSR, 24, 163–187.
Münkel, C., N. Eresmaa, J. Räsänen, and A. Karppinen, 2007: Retrieval of mixing height and dust concentration with lidar ceilometer. Bound.-Layer Meteor., 124, 117–128, https://doi.org/10.1007/s10546-006-9103-3.
Nappo, C. J., 2013: An Introduction to Atmospheric Gravity Waves. Academic Press, 359 pp.
NCEI, 2022: U.S. climate normals. NCEI, accessed 31 July 2022, https://www.ncei.noaa.gov/products/land-based-station/us-climate-normals.
NCEP, 2013: The GFS atmospheric model. NCEP Office Note 442, 14 pp.
Parajuli, S. P., and C. S. Zender, 2017: Connecting geomorphology to dust emission through high-resolution mapping of global land cover and sediment supply. Aeolian Res., 27, 47–65, https://doi.org/10.1016/j.aeolia.2017.06.002.
Parajuli, S. P., and C. S. Zender, 2018: Projected changes in dust emissions and regional air quality due to the shrinking Salton Sea. Aeolian Res., 33, 82–92, https://doi.org/10.1016/j.aeolia.2018.05.004.
Peckham, S. E., and Coauthors, 1991: WRF-Chem version 4.4 user’s guide. NCAR Tech. Rep., 68 pp., https://ruc.noaa.gov/wrf/wrf-chem/Users_guide.pdf.
Pokharel, A. K., M. L. Kaplan, and S. Fiedler, 2017: Subtropical dust storms and downslope wind events. J. Geophys. Res. Atmos., 122, 10 191–10 205, https://doi.org/10.1002/2017JD026942.
Poudel, U., S. Ahmad, and H. Stephen, 2021: Studying the intra-annual variability in surface area and volume of Salton Sea, California, using remote sensing-based water indices and GIS. World Environmental and Water Resources Congress 2021, Online, American Society of Civil Engineers, 769–783, https://doi.org/10.1061/9780784483466.070.
Prospero, J. M., P. Ginoux, O. Torres, S. E. Nicholson, and T. E. Gill, 2002: Environmental characterization of global sources of atmospheric soil dust identified with the Nimbus 7 Total Ozone Mapping Spectrometer (TOMS) absorbing aerosol product. Rev. Geophys., 40, 1002, https://doi.org/10.1029/2000RG000095.
Pu, B., and P. Ginoux, 2018: How reliable are CMIP5 models in simulating dust optical depth? Atmos. Chem. Phys., 18, 12 491–12 510, https://doi.org/10.5194/acp-18-12491-2018.
Ralph, F. M., P. J. Neiman, T. L. Keller, D. Levinson, and L. Fedor, 1997: Observations, simulations, and analysis of nonstationary trapped lee waves. J. Atmos. Sci., 54, 1308–1333, https://doi.org/10.1175/1520-0469(1997)054<1308:OSAAON>2.0.CO;2.
Ryerson, T., and Coauthors, 2013: The 2010 California Research at the Nexus of Air Quality and Climate Change (CalNex) field study. J. Geophys. Res. Atmos., 118, 5830–5866, https://doi.org/10.1002/jgrd.50331.
Scorer, R. S., 1949: Theory of waves in the lee of mountains. Quart. J. Roy. Meteor. Soc., 75, 41–56, https://doi.org/10.1002/qj.49707532308.
Shao, Y., and Coauthors, 2011: Dust cycle: An emerging core theme in Earth system science. Aeolian Res., 2, 181–204, https://doi.org/10.1016/j.aeolia.2011.02.001.
Skamarock, W. C., and Coauthors, 2019: A description of the Advanced Research WRF Model version 4. NCAR Tech. Note NCAR/TN-556+STR, 145 pp., https://doi.org/10.5065/1dfh-6p97.
Smith, R. B., 2019: 100 years of progress on mountain meteorology research. A Century of Progress in Atmospheric and Related Sciences: Celebrating the American Meteorological Society Centennial, Meteor. Monogr., No. 59, Amer. Meteor. Soc., https://doi.org/10.1175/AMSMONOGRAPHS-D-18-0022.1.
Stephen, M. F., and D. S. Gorsline, 1975: Sedimentary aspects of the new river delta, Salton Sea, Imperial County, California. Deltas: Models for Exploration, M. L. Broussard, Ed., Vol. 2, Houston Geological Society, 276–282.
Strauss, L., S. Serafin, and V. Grubišić, 2016: Atmospheric rotors and severe turbulence in a long deep valley. J. Atmos. Sci., 73, 1481–1506, https://doi.org/10.1175/JAS-D-15-0192.1.
Sweeney, M. R., E. V. McDonald, and V. Etyemezian, 2011: Quantifying dust emissions from desert landforms, eastern Mojave Desert, USA. Geomorphology, 135, 21–34, https://doi.org/10.1016/j.geomorph.2011.07.022.
Todd, M. C., R. Washington, S. Raghavan, G. Lizcano, and P. Knippertz, 2008: Regional model simulations of the Bodélé low-level jet of northern Chad during the Bodélé Dust Experiment (BoDEx 2005). J. Climate, 21, 995–1012, https://doi.org/10.1175/2007JCLI1766.1.
Wiegner, M., and Coauthors, 2014: What is the benefit of ceilometers for aerosol remote sensing? An answer from EARLINET. Atmos. Meas. Tech., 7, 1979–1997, https://doi.org/10.5194/amt-7-1979-2014.
Yang, S., J. Preißler, M. Wiegner, S. von Löwis, G. N. Petersen, M. M. Parks, and D. C. Finger, 2020: Monitoring dust events using Doppler lidar and ceilometer in Iceland. Atmosphere, 11, 1294, https://doi.org/10.3390/atmos11121294.
Yuan, T., S. Chen, J. Huang, X. Zhang, Y. Luo, X. Ma, and G. Zhang, 2019: Sensitivity of simulating a dust storm over central Asia to different dust schemes using the WRF-Chem model. Atmos. Environ., 207, 16–29, https://doi.org/10.1016/j.atmosenv.2019.03.014.
Zhao, A., C. L. Ryder, and L. J. Wilcox, 2022: How well do the CMIP6 models simulate dust aerosols? Atmos. Chem. Phys., 22, 2095–2119, https://doi.org/10.5194/acp-22-2095-2022.
Zucca, C., N. Middleton, U. Kang, and H. Liniger, 2021: Shrinking water bodies as hotspots of sand and dust storms: The role of land degradation and sustainable soil and water management. CATENA, 207, 105669, https://doi.org/10.1016/j.catena.2021.105669.