1. Introduction
Water is debatably our most crucial natural resource. Securing its supply in arid regions is a challenge, especially under increasing demand, and supply uncertainty in a globally warming climate. The glaciogenic seeding of clouds may help alleviate the shortage of water (Flossmann et al. 2019; Rauber et al. 2019). In a globally warming climate, precipitation and streamflow can change, as well as the efficacy of cloud seeding to increase precipitation.
There are many mountain ranges in the interior western United States with active glaciogenic cloud seeding programs (Fig. 1a). These programs mostly disperse silver iodide (AgI) nuclei released from generators on the ground or from aircraft. There are various modes of ice nucleation possible, including immersion, contact, condensation, and deposition freezing. The AgI nuclei typically contain a soluble part so they can act as cloud condensation nuclei, allowing for immersion-mode ice nucleation (Boe et al. 2014).
(a) Map of target areas of operational cloud seeding programs in the western United States in 2020 (source: North American Weather Modification Council; http://www.nawmc.org/). (b) Elevation map of the study area as represented in the 4-km-resolution IWUS model. Black dots mark SNOTEL sites. Beam paths of the three radiometers used for SLW evaluation are annotated with black (Boulder), gray (Cedar Creek), and red (Savery) arrows, with the radiometer located at the base of the arrow. The six numbered black lines show the locations of transects across mountain ranges of interest: 1) Glacier, 2) Payette, 3) Tetons, 4) Wind River Range, 5) Utah, and 6) Park and Rawah. The thick gray contours in (b) and all other maps below are U.S. state borders.
Citation: Journal of Applied Meteorology and Climatology 60, 9; 10.1175/JAMC-D-20-0288.1
Several recent field campaigns in the western United States have aimed to document the impact of glaciogenic seeding on clouds and precipitation, including the 2011–12 AgI Seeding Cloud Impact Investigation (ASCII) (Geerts et al. 2013; Pokharel and Geerts 2016), and the 2017 Seeded and Natural Orographic Wintertime Clouds: The Idaho Experiment (SNOWIE) (Tessendorf et al. 2019). These campaigns provided strong evidence that under the right conditions, glaciogenic cloud seeding from the ground (in ASCII) (Pokharel et al. 2017a, 2018, and references therein) and from an aircraft (in SNOWIE) (French et al. 2018; Friedrich et al. 2020) can enhance ice crystal production and surface precipitation. This evidence is based on observational case studies, mostly relatively shallow orographic clouds with only very light natural snowfall, accompanied by numerical simulations that capture the cloud microphysical impact of AgI seeding (e.g., Xue et al. 2013a; Geresdi et al. 2017). Observational evidence collected in ASCII, SNOWIE, and other field campaigns so far has been unable to narrow the range of conditions under which orographic clouds are seedable: for instance, the seeding efficacy of deeper clouds with heavier natural precipitation remains uncertain, for the simple reason that the primary seeding signature (enhanced radar reflectivity) is in the noise under stronger natural echoes, and the abundance of supercooled liquid water (SLW) in such clouds is inadequately measured.
The temperature and wind patterns favorable for the seeding of orographic clouds are quite well understood and predictable. For ground-released seeding material to be advected into orographic clouds, the flow must not be blocked. Orographic clouds are commonly targeted because of the frequent presence of SLW. Pockets of large SLW content are the result of an imbalance between water vapor flux (causing condensation) and ice particle growth. This is thought to happen often in updrafts, especially terrain induced ones. Clouds hugging the terrain may contain SLW, but the abundance and persistence of SLW is difficult to measure and poorly understood (Rauber et al. 2019; Flossmann et al. 2019).
Regional climate models with sufficient resolution to capture orographic flow and SLW patterns are increasingly being used to assess cloud seeding potential. For instance, Tessendorf et al. (2020) use a 4-km-resolution regional climate model to assess the feasibility of seeding specific target mountains in Wyoming. Pokharel et al. (2021) use the same model to examine the feasibility of orographic cloud seeding in Utah, both in the historical climate, and in a late twenty-first-century climate scenario. Both studies examine predefined target mountains. To our knowledge, such models have not been used to study the occurrence of seedable conditions on a broader scale, such as the entire interior western United States.
This study has two separate objectives: we examine (i) how frequently conditions are suitable for ground-based glaciogenic seeding across the entire interior western United States and (ii) how this is expected to change in a globally warming climate. This work does not quantify productivity (i.e., the amount of precipitation, snowpack, or streamflow attributable to seeding), although it does examine the amount of SLW present, which is a surrogate measure of productivity, nor does this study examine the economic feasibility of cloud seeding, that is, there is no cost–benefit analysis.
Section 2 describes the model used in this study, evaluates the modeled SLW, and defines “seedability.” Section 3 examines seedability in the current climate. Section 4 examines seedability changes in a globally warming climate. A discussion of the relevance, caveats, and nuances of this study follows in section 5. A summary of the key findings of this study is provided in section 6.
2. Regional climate simulations
a. Regional climate model
This study is based on output from the 30-yr interior western U.S. (IWUS) regional climate model (RCM) simulation (Wang et al. 2018). Details of the model architecture can be found in Wang et al. (2018), but two physics options should be mentioned here, because of their impact on the results presented: the Thompson cloud microphysics scheme (Thompson et al. 2008) and the YSU PBL scheme (Hong et al. 2006). Seedability is affected by the presence of SLW. The Thompson microphysics scheme, used in the IWUS simulations, was designed specifically to capture SLW, in particular for aviation safety (Thompson et al. 2008, 2017), and it captures SLW well (Merino et al. 2019; Xu et al. 2019). The PBL scheme is important as well in the study of ground-based seeding potential, as it impacts stratification, vertical mixing, and clouds in the boundary layer. The combination of the Thompson cloud scheme and the YSU PBL scheme captures orographic precipitation and the daily minimum and maximum temperature in the interior western United States very well in 4-km-resolution RCMs (Liu et al. 2011; Jing et al. 2017; Wang et al. 2018).
The IWUS domain encompasses the majority of the mountainous western United States excluding the West Coast. Figure 1b shows the analysis domain, which truncates the IWUS domain by 40 km on the south, west, and north boundaries, and by 105 km on the east boundary. IWUS has a 4-km grid spacing with 51 vertical levels, which appears to be sufficient to capture the complex flow and orographic precipitation patterns in this region (Ikeda et al. 2010; Liu et al. 2011).
IWUS was initialized and driven by the 6-hourly National Centers for Environmental Prediction Climate Forecast System Reanalysis (CFSR) (Saha et al. 2010). This study uses 10 years (2002–11) of cold-season data from the IWUS RCM. This period is referred to as the “historical climate.” The “cold season” is defined here as November–April, as ground-based seeding operations targeting orographic clouds in the western United States typically start sometimes in November, and end in April (e.g., Breed et al. 2014). Interannual variability is significant in the western United States, but the decade chosen for this analysis is representative of the 1981–2020 climatology, in terms of average cold-season precipitation and 700 mb relative humidity within the domain of interest (not shown).
For the “future climate” evaluations of seedability, the IWUS simulation was repeated to represent the climate around 2050, using a pseudo–global warming (PGW) technique (Jing et al. 2019). This approach follows Rasmussen et al. (2011, 2014) and Liu et al. (2017). The 30 years of PGW IWUS simulations assume the Intergovernmental Panel on Climate Change (IPCC) representative concentration pathway (RCP) 8.5 scenario. The difference, applied to the same CFSR driver dataset, equals the spatially resolved difference in phase 5 of the Coupled Model Intercomparison Project (CMIP5) RCP8.5 ensemble means between 2050 and 2000, on a monthly basis (Jing et al. 2019). The “future climate” in this study specifically refers to the last 10 years of the PGW IWUS output, that is, from 2052 to 2061. The RCP8.5 scenario, which is often referred to as the “business as usual” scenario in terms of human greenhouse gas emissions, may prove to exaggerate the rate of global warming (e.g., Hausfather and Peters 2020). In that case, the “future climate” conditions depicted here may apply only some decades later.
b. Evaluation of model supercooled liquid water
The IWUS simulations have been evaluated in terms of seasonal variation of precipitation and diurnal temperature range (Wang et al. 2018). The IWUS simulation appears to capture cold-season orographic precipitation more realistically than gauge-based gridded estimates such as the Parameter-Elevation Regressions on Independent Slopes Model (PRISM; Daly et al. 2008) over high terrain, specifically over the mountains where the gauge network [the Snowpack Telemetry (SNOTEL) network (Fig. 1b); Serreze et al. 1999] is sparse relative to the precipitation variability suggested by high-resolution models (Jing et al. 2017). This is not uncommon for high-resolution regional climate simulations of precipitation in complex terrain (Lundquist et al. 2019). IWUS’ good performance in terms of surface temperature and precipitation is a good reason to believe that the underlying dynamics and cloud processes are represented well as well.
Still, the cloud liquid water, which is critical in this study, has not been evaluated in the IWUS simulation. We use data from microwave radiometers stationed upstream of three mountain ranges in Wyoming (arrows in Fig. 1b) and operated during the study period (2002–11), although not for the entire time. Specifically, we use the Boulder radiometer data for 4.5 cold seasons (from November 2008 through December 2012), Cedar Creek data for one cold season (from November 2007 through April 2008), and Savery data for 3.5 cold seasons (from November 2009 through December 2012). The Boulder and Cedar Creek radiometers are 2-channel Radiometrics Corporation WVR-1100 units, and the Savery radiometer is a 5-channel Radiometrics WVP-1500 unit.
The radiometer data contain a drifting baseline, which results in an additive bias in the nonzero data. The baseline for the radiometer data at both Savery and Cedar Creek was corrected using asymmetric least squares smoothing (AsLS) (Eilers and Boelens 2005), in which the baseline is estimated iteratively with a smoothing function that weights positive deviations much less than negative ones. AsLS requires tuning of two parameters, the smoothness λ and the asymmetry p. These parameters were estimated manually for each season and each site separately and refined for subperiods if needed. As this method is very sensitive to spurious negative spikes, any such spikes were manually removed prior to baseline correction.
The 2-channel system has an accuracy (and noise level) of 25–30 g m−2 (Westwater et al. 2001; Crewell and Löhnert 2003; Turner 2007), and the 5-channel system’s accuracy is about 20 g m−2 (Crewell and Löhnert 2003). In all three cases, the radiometers were pointed at a grazing angle above the downwind terrain, in the directions shown in Fig. 1b. Generally, data from multiple elevation angles were available. We used the 8° elevation angle as it readily cleared the terrain at all three locations. The beam path (cone) of each of the radiometers was computed. The model LWC then was integrated along the corresponding slant path. Both the radiometer measurements and IWUS model data are expressed as a vertical liquid water path (LWP), units g m−2. The radiometers have varying time resolution of ~1 min, whereas the IWUS model output is hourly. Therefore, a 10-min averaging window around the top of each hour is used to compare with IWUS’ hourly snapshots. No time matching requirement is imposed on the model and radiometer distributions in Fig. 2. Since the cloud base temperature generally (>99% of the profiles) is below freezing during the cold season in our study domain, this comparison effectively evaluates model supercooled LWP (SLWP).
Distribution of LWP for three radiometers (along the beam paths shown in Fig. 1b)—(a) Boulder, (b) Cedar Creek, and (c) Savery (black)—and for time–space-matched IWUS model output (red). The numbers in the upper-right corner of (a)–(c) show the data record size as labeled, expressed in hourly samples exceeding the 25 g m−2 threshold.
Citation: Journal of Applied Meteorology and Climatology 60, 9; 10.1175/JAMC-D-20-0288.1
At first glance, the observed and modeled distributions of LWP are fairly similar, with small model biases for the three different locations (Figs. 2a–c). To assess similarity between the two distributions, we used a two-sided two-sample Kolmogorov–Smirnov test and an unpaired version of the Wilcoxon test. Both tests indicate high confidence (at least 99% certainty) that the distributions are different, at all three sites. This is due to a consistent difference in the frequency of very low LWP occurrences, which appears to be underestimated by the model, resulting in an underestimate of the frequency of SLW occurrence by almost a factor of 2, on average at the three sites. But that may be a measurement uncertainty. In Fig. 2, a (10-min average) 25 g m−2 noise floor is assumed for all three radiometers, consistent with recommendations in the abovementioned literature. A higher LWP threshold, for example, 40 g m−2, results in a better correspondence in event frequency. Higher LWP values become increasingly rare in these cold conditions, however, thereby reducing the statistical significance of the comparison. Unfortunately, an effort to obtain seasonal data records from more radiometers within the IWUS domain was not successful. In any event, assuming a 25 g m−2 noise floor, radiometer data indicate that SLW occurs 7%–10% of the time at these three locations in the cold season, whereas the IWUS model suggests a lower frequency (4%–7%). This frequency underestimation may be due to the coarse model terrain resolution.
In summary, the IWUS appears to underestimate the frequency of low SLWP values, in the 25–50 g m−2 range (Figs. 2a–c), and thus the fraction of time conditions that are suitable for seeding (defined below) may be underestimated as well. On the other hand, the IWUS may slightly overestimate SLWP when conditions are suitable. These two statements have a low confidence level, since most LWP values are close to the radiometer noise floor.
c. Criteria for seedability
1) Maximum mixing depth
This section describes the criteria deemed necessary for effective ground-based glaciogenic seeding. Ground-based AgI dispersal typically occurs from towers 2–10 m tall by atomizing a seeding solution into a propane flame producing microscopic AgI crystals (e.g., Breed et al. 2014). The seeding material then is mixed in the boundary layer (BL) flow. Therefore, we examine the flow within the model’s output variable planetary boundary layer (PBL) depth. We assume here that the AgI will be able to mix vertically throughout the PBL, and that some detrainment of PBL air into the free troposphere occurs. This is common in complex terrain, for instance, in breaking waves, BL separations over crests, and hydraulic jumps. Therefore, the flow within the model PBL height plus an additional isentropic depth of 4 K above the PBL is examined. Here, this depth is referred to as the maximum mixing depth (MMD; m AGL). The additional height above the PBL is measured in potential temperature instead of distance so that the depth of assumed mixing is dependent on PBL capping strength: in a well-capped BL, which detrains little, the 4 K extra adds very little physical depth. In an uncapped BL, with more exchange into the overlying free troposphere, the MMD is considerably deeper than the PBL depth. The choice of the PBL depth plus 4 K is arbitrary, and further work is needed to examine actual dispersal of AgI nuclei near complex terrain using very high-resolution simulations (as in Chu et al. 2017a,b). For reference, the MMD is 1.1 km deeper than the PBL depth in the standard atmosphere, where potential temperature increases by 3.5 K km−1. We realize ground-released AgI nuclei will not mix throughout the MMD in a highly stratified BL, but in that case conditions will likely be flagged as unseedable because of blocked flow, as discussed below.
2) Flow blocking
Typically, ground generators are placed in the foothills of mountain ranges (target regions) on the windward side of the mountains. The AgI-bearing air parcels must reach water vapor saturation to be effective. Therefore, the flow must be uninhibited from flowing up and over the mountains closest to any given location. In general, the higher in the foothills and the closer to the downwind mountain crest, the less likely the BL flow is blocked (Espitalié 2020) but the shorter the distance of potential AgI impact on cloud processes and precipitation; such impact ceases with the subsidence-induced evaporation of cloud droplets in the lee of the crest. Thus, the optimal location for ground generators is somewhere between the lower foothills (where blocking is a challenge) and the upper foothills (where the remaining in-cloud trajectory length becomes short).
To continuously evaluate the degree of flow blocking in a complex terrain environment, a technique is developed to assess blocking intensity relative to downstream terrain locally, at any grid point. The bulk Froude number is a widely used metric for this purpose, typically measured at a specific, arbitrarily selected point upwind of a specific mountain range. Traditionally it is expressed as Fr = U/(NH), where U is the layer-mean horizontal wind component parallel to the terrain elevation gradient, N is the layer-mean Brunt–Väisälä frequency, and H is the height of the target mountain range. The bulk Froude number is computed over a depth corresponding with the average height of the downwind terrain, building on numerous modeling studies using idealized terrain (Markowski and Richardson 2011, chapter 13). Over real terrain, this height and the range’s orientation must be subjectively chosen. Here, Fr is computed locally, for the entire analysis domain, alleviating the need to define parameters a priori, as follows.
The bulk Froude number is computed over the local MMD, defined above. Here, U is defined as the mean wind along the terrain slope between the ground and the MMD. In the interior western United States, the local terrain slope generally points toward the more significant terrain because, at 4-km grid spacing, fine undulations in slope are not resolved. The MMD-average Brunt–Väisälä frequency N is the square root of the vertically averaged N2 values. The dry N2 (based on virtual potential temperature) is used below the lifted condensation level (LCL) of the surface parcel. Above this LCL, the moist N2 is computed [Markowski and Richardson (2011), their Eq. (3.9)] for each model level within the MMD. Here, we use equivalent potential temperature as in Eq. (39) of Bolton (1980), and the moist adiabatic lapse rate and temperature-dependent latent heat of vaporization from Yau and Rogers (1996). The impact of moisture on stability is important to include because the times and areas of interest often have a relatively low cloud base. Because we are concerned with low-level stability, we use the LCL as the boundary between dry and moist N2 values. This approach captures changing stability upon mountain approach.
For any given point in the domain, the closest terrain crest, the direction toward it, and the height of that crest need to be defined a priori. This is accomplished by first finding the local gradient in terrain elevation. The direction of this gradient is used to compute the component of the wind flowing toward the barrier. In this calculation, the time-dependent U wind is negative if the wind is blowing “downslope.” Next, the difference between the local elevation and highest terrain in the uphill direction within 200 km is used as H in the Fr calculation. The distance of 200 km is chosen based on examining the Rossby radius of deformation (defined here as Lr = N × MMD/f) during an entire cold season; generally, Lr < 200 km in the region and at the times of interest. Using this upper bound maximizes the H in Froude calculation, thus minimizing Froude number. Rossby radius of deformation here is interpreted as the distance over which downwind blocking terrain can influence the flow (Markowski and Richardson 2011, p. 347).
With this specific technique for calculating the time-dependent Froude number at every location in a model domain, Fr values can be grouped in four distinct classes: 0 < Fr < 0.5 represent considerably blocked flow, 0.5 < Fr < 1.0 represent marginally blocked flow, and Fr > 1.0 represents unblocked upslope flow. Unlike the traditional approach, we also consider Fr < 0 (downslope flow). When Fr is undefined (neutral or statically unstable conditions), the airflow can move over the mountain unimpeded and possibly release convection if the flow is upslope. Therefore, this condition is included in the unblocked upslope category. Conditions with Fr undefined and downslope flow are included in the Fr < 0 group (downslope flow).
3) Seedability conditions
Three local criteria are used to determine whether a specific time is deemed seedable at a specific location: 1) unblocked upslope flow (Fr > 1.0, or upslope statically neutral/unstable conditions), 2) temperature at or below −6°C somewhere within the MMD, and 3) SLW is present, that is, at least 10 g m−2 LWP at or below −6°C within the MMD.
For criterion 2, the −6°C temperature threshold is commonly used (Rauber et al. 2019, and references therein), since the nucleation of ice in a supercooled liquid cloud through immersion or contact freezing by AgI nuclei requires temperatures at least this low (DeMott et al. 1997). Unlike other studies (e.g., Tessendorf et al. 2020), no lower temperature bound is used in this study, simply because liquid water becomes exceedingly rare at temperatures below about −20°C in the IWUS model data, and in reality. For criterion 3, Pokharel et al. (2021) use the same LWP threshold (10 g m−2), but at any temperature below freezing.
The one drawback of this local definition is that it does not account for transport from the source (the AgI generator on the ground) into cloud, and further transport of the embryonic ice crystals as they grow and eventually precipitate on the ground. Such transport often exceeds the grid spacing of the model used here (4 km), and hence is not local. Accounting for this transport requires a Lagrangian approach, which is beyond the scope of this study.
4) Regions of interest
Map depicting the four mountainous regions examined below: NI is Northwest Interior, GY is Greater Yellowstone, UR is Utah Rockies, and CR is Colorado Rockies. East- and west-facing slopes are delineated in dark and light shading, respectively. The color scheme used here remains consistent for later figures. The thin gray contours on this map and on all other maps below show terrain elevation, at 500-m intervals.
Citation: Journal of Applied Meteorology and Climatology 60, 9; 10.1175/JAMC-D-20-0288.1
3. Seedability in the current climate
In this section, the potential for ground-based glaciogenic seeding in the current climate, or rather, the “historical climate” (i.e., the 2002–11 time period), is examined. First, the cold-season (November–April) climatology is characterized (Fig. 4). The MMD generally is < 1.5 km over the western slopes in the cold season, but deeper on the east slopes of the mountain ranges facing the High Plains (Fig. 4a). This may be the result of frequent lee flow separation and downslope (“Chinook”) flow (e.g., Pokharel et al. 2017b). Temperature (Fig. 4b) is below freezing not just over the high terrain, but also in high valleys and basins in a continuous belt from southern Colorado to northwest Montana. Dewpoints are highest in the northwest and taper off to the south (Fig. 4c). Subsidence in the lee of the easternmost ranges leads to much lower relative humidity (RH) (Fig. 4d).
Maps of November–April 2002–12 average conditions, vertically averaged within the MMD: (a) average cold-season MMD (m AGL), (b) temperature (K), (c) dewpoint (K), and (d) relative humidity (%).
Citation: Journal of Applied Meteorology and Climatology 60, 9; 10.1175/JAMC-D-20-0288.1
a. Spatial distribution of seedability and seeding efficacy
The mountains are the most seedable places in any given region, according to the three criteria listed in section 2c (Fig. 5). The most seedable mountains are around Glacier National Park, Montana, and in the Wind River Range in Wyoming, where the peak seedability values approach 40% of the time in the cold season. A value of 40% corresponds with 72 of ~180 full days of seeding.
Percentage of time that conditions are deemed suitable for ground-based glaciogenic cloud seeding during the cold season (November–April), in the historical climate.
Citation: Journal of Applied Meteorology and Climatology 60, 9; 10.1175/JAMC-D-20-0288.1
To explain seedable frequency (Fig. 5), we examine the percentage of time the three criteria separately are met (Fig. 6). The temperature criterion is almost always satisfied in the highest elevations during the cold season (Fig. 6a). Unblocked upslope flow is very common over the western slopes of mountain ranges, up to ~80% of the time (Fig. 6d). Unblocked upslope flow is considerably less common on east-facing slopes than their west-facing counterparts, because of the predominantly westerly flow during the cold season. SLW presence within the MMD is much less common, reaching up to ~40% of the time over the high mountains in the northwest interior and greater Yellowstone (Fig. 6b). Because the SLW criterion is a subset of the temperature criterion, it is deduced that the liquid water presence is the limiting factor to seedability.
Percentage of time the MMD has (a) temperatures ≤ −6°C and (b) SLW at temperatures ≤ −6°C, (c) MMD-average SLWP ≤ 0°C (g m−2) while seedable, (d) percentage of time with unblocked upslope flow.
Citation: Journal of Applied Meteorology and Climatology 60, 9; 10.1175/JAMC-D-20-0288.1
The distribution of SLW frequency in the lower troposphere (the MMD) is much different from the distribution of SLW-topped cloud frequency, as seen by MODIS, a spaceborne sensor (Morrison et al. 2013). SLW-topped clouds are most common in the intramountain basins such as the Snake River Plain in Idaho (Fig. 4 in Morrison et al. 2013), while MMD-based SLW is closely associated with the terrain (Fig. 6b). The implication is that, in general, satellite-based guidance cannot be used for ground-based seeding operations, although progress in LWP estimation from space is being made in very limited situations (e.g., Leinonen et al. 2016).
Given that up to 40% of the winter season can be seedable (Fig. 5), it raises the question of how productive this seeding could be. The average SLWP (i.e., at any temperature < 0°C) within the MMD during ground-based seedable times (Fig. 5) is mapped in Fig. 6c. First, it should be noted that, during seedable times, the MMD is close to the overall average (Fig. 4a) over mountains, where in the intramountain basins, it is much higher (not shown). (Seedable conditions are satisfied in basins only in the rare occasions of deep frontal disturbances with moist-adiabatic lapse rates, and thus a very deep MMD.) In general, seeding efficacy (i.e., the seeding-induced additional precipitation on the ground) is proportional to the amount of cloud SLW (Li and Pitter 1997; Xue et al. 2013a; Geresdi et al. 2017), so Fig. 6c is a good measure of seeding efficacy. In the assessment of operational cloud seeding feasibility, both the frequency of seedable conditions (Fig. 5) and the seeding efficacy (when clouds are seedable) (Fig. 6c) must be considered. A temperature threshold of −6°C is used as criterion for seedability because of AgI nucleation properties (section 2c). But when seedable conditions prevail, all of the LW colder than 0°C can contribute to productivity, not just the LW below −6°C. This is why Fig. 6c (and others below) examine the full SLWP. In Wyoming, seeding efficacy is considerably higher in the western mountains (including the Teton and Wind River Ranges) than in the Bighorn Range to the east (Fig. 6c), consistent with Tessendorf et al. (2020, their Fig. 18a).
Areas with considerable seedability (>20% of the time) are far more common in the two northern mountain regions (NI and GY), while areas with poor seedability (<10% of the time) are proportionally more dominant in the two southern mountain regions (CR and also UR) (Fig. 7a). Our proxy to seeding efficacy, that is, the SLWP when seedable, shows different distributions in the four regions of interest (Fig. 7b): promising SLWP values (>100 g m−2) are most common in the two western regions (NI and especially UR), while low SLWP values (<100 g m−2) are relatively more common in CR and especially in GY. (Areal averages are summarized in Fig. 8.) In general, NI is most frequently seedable, and CR is the least. The seeding efficacy (cloud SLWP) is lowest in GY, and highest in UR, on average. It is higher still over isolated mountain ranges of northern Arizona (Fig. 6c), where seedable conditions are exceedingly rare (Fig. 5). In general, seedable frequency and efficacy are 2–3 times as high over the western slopes as over their eastern counterparts (Fig. 8).
Cumulative distribution function of area of (a) seedability and (b) seeding efficacy in the four mountain regions of interest, in the historical and future climates. The ordinate is the area normalized by the total area of each region.
Citation: Journal of Applied Meteorology and Climatology 60, 9; 10.1175/JAMC-D-20-0288.1
Regional average seedable frequencies (Fig. 5) and SLWP values during these times (Fig. 6c) during the cold season in the historical and future climates. The columns are color coded by mountain region, consistent with Fig. 3.
Citation: Journal of Applied Meteorology and Climatology 60, 9; 10.1175/JAMC-D-20-0288.1
We now examine seedability variations across select mountain ranges using transects (Fig. 9). Both the seedability (fraction of time seedable) and seeding efficacy are examined. Both parameters peak some distance (~4–6 km) upwind (i.e., to the west or southwest) of mountain ridges. This applies in particular to the relatively isolated ranges of the Wind River Range (Fig. 9c), the Tetons (Fig. 9b), and the Utah range (Fig. 9d). Also, both parameters decay more rapidly to the east (generally downstream) than to the west (generally upstream) of crests. Multiridge mountain complexes such as Glacier (Fig. 9f) and the Payette (Fig. 9a) generally have the highest SLW near the (south)westernmost ridge and decreasing amounts farther downwind, even though the downwind ridges may be higher. These downwind ridges are generally as frequently seedable as the upstream ridges, but with less cloud SLW, the seeding may be less productive (Fig. 9). The Glacier and Park–Rawah transects extend into the High Plains, and it can be argued that the northeast side of the Wind River Range transect also connects to the High Plains (Fig. 1b). The foothills facing the High Plains typically are very dry (Fig. 4d), and their cold-season precipitation mostly derives from “upslope” storms (Boatman and Reinking 1984). Such storms in this region (including the east sides of the Glacier, the Park–Rawah, and Wind River Range transects) are most common in spring, generally peaking in the month of April, which is included in this analysis. Yet there is no secondary maximum in the frequency or intensity of seedable conditions over the eastern foothills for these three transects: at best these transects reveal a “shoulder” in seedability just east of the easternmost peak (Fig. 9). The lack of a maximum appears to be due to the relatively high minimum temperature within the MMD in April, usually above −6°C (not shown). Upslope storms often contain convection carrying BL air to lower temperatures above the MMD, and thus they may be seedable, but that is beyond the scope of this study.
Transects of terrain height (solid colored lines) from west (or southwest) to east (or northeast) along the prevailing cold-season wind for six mountain ranges (transect locations annotated on Fig. 1b), and variation of four parameters along this terrain. These four parameters are, from top to bottom in each panel, percentage of time seedable (gray-tinted bar), its change in the future climate (diverging red- and blue-tinted bar), average SLWP within the MMD during seedable times (blue-tinted bar), and its change in the future climate (diverging red- and blue-tinted bar). The color key bars on the right follow the same sequence. The abscissa is distance from the crest or, if multiple ridges are present, distance from the center of the mountain range.
Citation: Journal of Applied Meteorology and Climatology 60, 9; 10.1175/JAMC-D-20-0288.1
To better understand the distribution of flow blocking, we examine the distribution of Fr in the various mountain ranges (Fig. 3) for the entire study period (Fig. 10). The curves in Fig. 10 are normalized by the total frequency, so if the graph were extended to −∞ < Fr < ∞, the area under the various curves would be equal to 100. Unblocked upslope flow (Fr > 1) is most common on the western slopes of all the ranges rather than the eastern. Blocking is a fairly complex phenomenon, impacted by the specific geometry of the mountains as well as westerly flow and stability of upstream environments. Eastern slopes experience downslope flow (Fr < 0) most of the time.
Normalized distribution of local Froude number in each of the regions of Fig. 3. A negative Fr indicates downslope flow. The inserted table lists the areas under the curve in four bins of Fr values.
Citation: Journal of Applied Meteorology and Climatology 60, 9; 10.1175/JAMC-D-20-0288.1
b. Number and duration of cases
To further understand the nature of these seedable conditions, it is useful to examine how long they persist and how many seedable “cases” there are in a season, in addition to the cumulative amount of time seedable in a season discussed above. Here, cases can have a duration and separation as little as one hour, the time resolution of the IWUS dataset. Attempts to group events together by a threshold of separation time was explored but avoided, because it reduces the time resolution and straightforward interpretation of the analysis.
In each of the various mountain ranges there are between 20 and 40 cases per year that persist for no longer than about 12 h (Fig. 11). The distribution of intercase gaps (separation time between cases) peaks at the minimum (1 h) and drops rapidly with length (not shown), so many storm systems contain multiple short seedable periods. The number of cases is largest in the GY region and smallest in UR (Fig. 11a). Cases last longest in NI and are most transient in UR (Fig. 11b). Again, we decompose case duration into its three criteria, to understand the patterns observed (Fig. 12). The temperature criterion is the most persistent, while the SLW criterion is the most transient, showing, again, that the SLW is the limiting factor. Note that the steepest part of the cumulative distributions in Figs. 11 and 12 corresponds with the mode of the distribution. Therefore, most seeding cases are only an hour long (Fig. 11b). Most periods of cold-enough temperatures are 10 h long, periods with SLW mostly last 2 h, and periods with unblocked upslope flow mostly last 3 h (Figs. 12d–f). Seedability is a question of how these three criteria coexist, rather than the presence of each one individually.
(a) The number of periods with seedable conditions (“cases”) per cold season for ground-based cloud seeding in each of the four regions (east- and west-facing slopes) during the study period (2002–11) (solid), and its change into the future climate (hatched). The black or gray dot respectively represents the actual average number of seeding cases per year in an operational program on the southwest or northeast side of the Wind River Range in Wyoming from 2007 to 2017. (b) Cumulative distribution of the duration of the cases from (a) for each range (2001–11 cold seasons) (solid lines), and the change in the cumulative distribution into the future climate (dashed lines). Black or gray dots respectively represent the cumulative frequency distribution of duration of the actual seeding cases in the operational program on the west or east side of the Wind River Range.
Citation: Journal of Applied Meteorology and Climatology 60, 9; 10.1175/JAMC-D-20-0288.1
As in Fig. 11, but separated into the three seedability criteria: (a),(d) temperature ≤ −6°C; (b),(e) SLW presence at temperatures ≤ −6°C; and (c),(f) unblocked upslope flow, all within the MMD.
Citation: Journal of Applied Meteorology and Climatology 60, 9; 10.1175/JAMC-D-20-0288.1
In practice, operational ground-based seeding events are fewer, but last much longer. This is illustrated for 10 years of seeding performed in the Wind River Range (Fig. 11, black and gray dots). This large discrepancy is largely explained by the “when in doubt, seed” approach that the operations in the Wind River Range and elsewhere generally take. Model guidance is improving but remains uncertain, even now that many seeding operations are guided by customized, highly resolved numerical weather prediction models. Even if the model guidance was perfect, operators would tend to seed longer than suggested by our seedability criteria, bridging gaps and covering shoulder periods. Most model-defined seeding cases are very short and separable by as little as one hour. This does not fit with the complexities of operating AgI generators. It also neglects the last step, that is, AgI dispersal across the target area and seeding-induced cloud and precipitation process, which themselves often take more than one hour. Airborne programs have a slight advantage in this regard, they can actively and directly observe SLW, and only seed clouds that are observed to be seedable, while depending on model and radiometer guidance for flight periods.
4. Change in seedability in a warmer climate
a. Future climate scenario
We now examine whether seedability and seeding efficacy will change in a warmer climate. Figure 13 gives some general insight into the change in the environment in the future climate assumed in the IWUS PGW simulation. As mentioned in section 2b, this climate is believed to be representative of conditions anticipated by the mid- to late twenty-first century. On average the MMD only changes by a relatively small amount (<50 m) in either direction, considering the MMD is between 1.0 and 1.6 km (Fig. 4a). The MMD generally decreases over the tallest mountains and eastern slopes, while it increases in the interior basins, consistent with the projected reduction in snow cover there in a warmer climate (not shown). The main change is the temperature, predicted to be 2.0 K warmer on average in the MDD during the cold season, according to the IWUS PGW simulation. It is important to keep in mind that this PGW technique is performed by perturbing the driver reanalysis dataset (CFSR) in a dynamically consistent way, implying an identical sequence of weather events, in a perturbed climate. Some high-elevation basins see a more pronounced warming (Fig. 13b), on account of a significant reduction in snow cover, that is, a snow–albedo–temperature feedback (e.g., Minder et al. 2016). The absolute humidity (dewpoint) in the MMD increases everywhere, by 1.8 K on average in the domain, the larger increases found mostly on the east side of the Continental Divide (Fig. 13c). As a result, the relative humidity is slightly lower on the west side of the divide in the future climate, relative to the historical climate, and slightly higher on the eastern slopes. This is related to the average winds becoming less predominantly westerly in the future climate: the IWUS simulations indicate an increase in the frequency and/or intensity of upslope flow events on the eastern slopes in winter (Jing et al. 2019). The confidence level of these climate change predictions is beyond the scope of this study. The CMIP ensemble mean guidance is commonly used as the best available for climate impact studies. Here, we merely examine the impact of this change on cloud seeding potential.
As in Fig. 4, but showing the absolute difference between the future and the historical climates.
Citation: Journal of Applied Meteorology and Climatology 60, 9; 10.1175/JAMC-D-20-0288.1
b. Changes in seedability
All four mountain regions of interest (Fig. 3) retain fewer pockets with good seedability. Areas seedable >20% of the time become exceedingly small and survive virtually only in the two northern regions (Fig. 7a). Almost the entire interior western United States sees a decrease in seedability from the historical climate to the future climate in the cold season (Fig. 14). The majority of the domain sees a small decrease in the percentage of time seedable, because this percentage was small to begin (Fig. 5). Of the four mountain regions of interest (Fig. 3), the NI region sees the largest decrease on average, from 10.3% to 9.1% of the time seedable (Fig. 8), while the decrease is the smallest in CR. The spatial patterns of seedability (Fig. 5) and seedability (Fig. 14) change in Utah are very similar to those predicted by Pokharel et al. (2021, their Fig. 6), with the Uinta mountains being the most seedable, with little or no decrease in a warmer climate. Since Pokharel et al. (2021) use a different regional climate model, different climate change guidance, and a slightly different definition of seedability, this correspondence points to the robustness of these findings.
As in Fig. 5, but showing the absolute change in percentage of time seedable in the future climate.
Citation: Journal of Applied Meteorology and Climatology 60, 9; 10.1175/JAMC-D-20-0288.1
Some locations in the domain, mainly the eastern foothills of the larger ranges in the CR and GY regions, are expected to become slightly more seedable (Fig. 14). This is evident in the transects for the Wind River Range (Fig. 9c) and Park–Rawah (Fig. 9e) also. This can be explained by the increase in relative humidity on these eastern slopes (Fig. 13d) in the future climate.
To explain the change in seedability, we examine the changes in the three criteria (Fig. 15). First, cold-enough temperatures become much less likely in the MMD, and the greatest absolute decreases in frequency, up to 14% of the time, are on the slopes of the mountains (Fig. 15a). This may be explained in part by the fact that the mean warming (from the historical to the future climate) is larger on mountain slopes where the temperature profile is closer to the moist adiabatic than the dry adiabatic condition (Xue et al. 2020). The southernmost mountains see the largest decrease in frequency of cold-enough temperatures over the mountain peaks (Fig. 15a).
As in Fig. 6, but showing the absolute difference between the future and the historical climates.
Citation: Journal of Applied Meteorology and Climatology 60, 9; 10.1175/JAMC-D-20-0288.1
Second, LW at temperatures of <−6°C also becomes less common, especially in NI (Fig. 15b), with decreases that approach the absolute frequency in the historical climate (Fig. 6b) in some places, for example, at the lower elevations in north Idaho and northeast Washington. The slight increase in seedability over the eastern foothills of the larger ranges in CR and GY regions (Fig. 14) can be explained by the increase in frequency of SLW presence there (Fig. 15b), which in turn is related to the increase in RH (Fig. 13d) and the changing wind patterns.
Third, the changes in frequency of unblocked upslope flow (Fig. 15d) are smaller and more localized than the two other criteria. In general, the eastern mountain slopes see increases, while the western slopes see decreases, especially in the GY area, simply because easterly upslope flow becomes more common.
Both the number and duration of “cases” (defined in section 3b) is expected to decrease, with the largest (smallest) decreases in NI (CR) (Fig. 11). This results from warming and lack of SLW colder than −6°C: in terms of the temperature criterion, there is a significant decrease in the duration of cases (Fig. 12d), and some cases in the historical climate split up in two shorter cases in the future climate, resulting in a small increase in number of cases (Fig. 12a). The number and the duration of cases with SLW decreases (Fig. 12b). The number and duration of unblocked upslope flow cases hardly changes (Figs. 12c,f). Interestingly, the largest changes in the durations of these cases (Figs. 11b, 12d–f) are for the mode of durations, that is, where the CDF is steepest.
c. Changes in seeding efficacy
The efficacy of ground-based seeding is expected to increase in a warmer climate (Fig. 15c): in most areas there will be more SLW available under seedable conditions, especially in the northwest interior, and over the mountains of northern Arizona and southern Utah. In a relative sense, the seeding efficacy is expected to increase by 10%–15% in the four mountain regions (Fig. 8). All four mountain regions of interest see significant increases in frequency of high seeding efficacy (>100 g m−2) (Fig. 7b). In general, the eastern slopes of the four mountain regions of interest see a larger increase in seeding efficacy, yet a smaller decrease in seedable frequency (Fig. 8), so the seeding of eastern slopes become relatively more attractive in a warmer climate. Seeding operations conducted since 2007 over the Wind River Range in Wyoming, for instance, employ nine seed generators on the southwest side of the range, and one on the northeast side (Tessendorf et al. 2020).
5. Discussion
Here, some nuances, caveats, and ramifications of the findings of this study are addressed. First, Figs. 5 and 6c show where seedable clouds frequent, but they do not provide guidance for the exact location of AgI generators. In practice, AgI generators should be placed some distance upwind of the most seedable areas, to allow the seed material to mix well in the PBL to allow ice initiation, snow growth, and eventually snowfall onto the ground. Stronger winds tend to deepen PBL mixing of ground-released nuclei, but they shorten the advection time, which should be long in comparison with the microphysical growth and fallout time scales. This study does not address the question of optimal placement of AgI generators relative to the target clouds. Optimizing generator placement calls for more resolved dispersion, Lagrangian, and/or AgI-resolving modeling. There often is a trade-off: too far upstream may result in poor targeting or reduced PBL mixing and no advection over the target mountain, on account of frequently blocked flow; too close to the crest may result in insufficient in-cloud residence time. A favorable generator position in one storm may be a bad one in another.
Second, it has been shown that SLW is the limiting factor in the seedability climatology, but it is also the most uncertain one. The IWUS appears to underestimate the frequency of low SLWP values (suggesting that the actual seedability could be slightly larger than presented here), but the IWUS may overestimate seeding efficacy when conditions are suitable. Although the Thompson microphysics scheme used herein is considered as the preferred microphysics representation for this purpose because of its ability to represent SLW accurately, SLW content is highly uncertain and different models’ microphysics have different SLW contents (e.g., Tan et al. 2016; Lenaerts et al. 2017). Because seedability depends mainly on SLW presence, the results are sensitive to the model physics choices made, and this sensitivity is not quantified in this study. Even if it were quantified with a number of different cloud schemes, each over at least 10 years in the present and future climates, uncertainty would remain, since SLW is highly dependent on natural concentrations of ice nucleating particles (e.g., DeMott et al. 2016; Solomon et al. 2018) and on secondary ice production processes, which remain poorly understood (e.g., Field et al. 2017). If ice nucleating particles are naturally abundant (such as mineral dust), orographic precipitation is naturally more efficient and SLW less abundant, therefore rendering AgI seeding in these situations less useful (e.g., Muhlbauer and Lohmann 2009).
Third, this study has been specifically tuned to the commonly used ground-based dispersal of AgI nuclei. Airborne seeding is becoming more popular in operational orographic cold-season seeding programs, and this study is not tuned to that dispersal technique. In essence, airborne seedability and seeding efficacy, using in situ or ejectable flares, only depends on the presence and abundance of SLW (both below −6°C and at any temperature) at any level, not just within the MMD (e.g., Tessendorf et al. 2019, 2020). A similar climatology of airborne seedability, based on regional climate model output, is needed. Our work can be readily extended to airborne seeding by omitting the requirement that the low-level flow must be unblocked. To quantify uncertainty, we recommend that an ensemble approach is assumed for this, using different model physics, especially different cloud microphysical schemes.
Last, the question of seeding efficacy is addressed only superficially, as it was assumed that the SLW path is a good proxy for potential precipitation enhancement. Ultimately, the climatology of potential seeding impact would be addressed best with a model that captures the full cloud microphysical impact of AgI injection (e.g., Xue et al. 2013b; Rasmussen et al. 2018). Such approach can also be used to systematically seek the best location of AgI generators, over the course of multiple cold seasons. Such simulations are computationally demanding. Other potential follow-up work that could provide useful is (i) to do sensitivity testing on cloud seeding criteria used here, for example, the definition of MMD, or unblocked flow, and (ii) to correlate these seedable conditions to synoptic scale weather patterns for forecasting purposes.
6. Conclusions
This study uses 10 years of a high-resolution regional climate simulations of the interior western United States to address ground-based glaciogenic seedability in the historical climate (2002–11), and in a future climate that is about 2 K warmer than the historical climate, a climate that will be reached by ~2050–60, according to the IPCC RCP8.5 scenario. The main conclusions of this study are as follows:
A comparison with seasonal radiometer data at three locations indicates that the IWUS (which uses the Thompson microphysics scheme) captures the frequency and amount of supercooled liquid water reasonably well. Uncertainty exists about the model frequency of low SLWP values, in the 10–50 g m−2 range, because this range of values is around the radiometer noise floor.
The highest mountain ranges, especially those in the northwest interior and the greater Yellowstone area, are generally most often seedable, up to 20%–30% of the time during the cold season (November–April), according to the IWUS. Both the frequency of suitable conditions (“seedability”) and the abundance of SLW (a proxy for seeding efficacy) peaks just west of mountain crests, because of the prevailing westerly flow. In general, orographic clouds over the western ranges are more seedable and carry more SLW than those farther east, especially those east of the Continental Divide.
A comparison between seedable areas and current operational seeding programs (Fig. 1a) suggests that a considerable area of the interior west has the potential for seeding programs, in particular in less arid mountain ranges in the northwest interior.
Seedable mountain ranges in central/southern Utah and northern Arizona are relatively small, and they are rarely seedable, but, when seedable, the SLW path is relatively high.
In a warmer climate, seedability is expected to decrease over almost the entire interior western United States. Island areas that are highly seedable (>20% of the time) become exceedingly small and rare. Yet, during suitable time periods, the seeding efficacy is expected to be slightly higher (10%–15% on average) everywhere. The seedability decrease is largest in the most seedable areas, such as the northwest interior, yet the more arid eastern slopes of the main mountain regions in the interior west may see a small increase in seeding efficacy.
Acknowledgments
This work was funded by the Wyoming Water Development Commission and the U.S. Geological Survey, under the auspices of the University of Wyoming Water Research Program. The National Center for Atmospheric Research (NCAR) is sponsored by the National Science Foundation. The numerical simulations were conducted on Yellowstone and analysis was done on Cheyenne at the NCAR Wyoming Supercomputer Center using Grant WYOM0070.
Data availability statement
The IWUS model output, for the historical climate, is available from https://doi.org/10.5281/zenodo.1157112. The future climate is available from https://doi.org/10.5281/zenodo.3934896.
REFERENCES
Boatman, J. F., and R. F. Reinking, 1984: Synoptic and mesoscale circulations and precipitation mechanisms in shallow upslope storms over the western High Plains. Mon. Wea. Rev., 112, 1725–1744, https://doi.org/10.1175/1520-0493(1984)112<1725:SAMCAP>2.0.CO;2.
Boe, B. A., J. A. Heimbach, T. W. Krauss, L. Xue, X. Chu, and J. T. McPartland, 2014: The dispersion of silver iodide particles from ground-based generators over complex terrain. Part I: Observations with acoustic ice nucleus counters. J. Appl. Meteor. Climatol., 53, 1325–1341, https://doi.org/10.1175/JAMC-D-13-0240.1.
Bolton, D., 1980: The computation of equivalent potential temperature. Mon. Wea. Rev., 108, 1046–1053, https://doi.org/10.1175/1520-0493(1980)108<1046:TCOEPT>2.0.CO;2.
Breed, D., R. Rasmussen, C. Weeks, B. Boe, and T. Deshler, 2014: Evaluating winter orographic cloud seeding: Design of the Wyoming Weather Modification Pilot Project (WWMPP). J. Appl. Meteor. Climatol., 53, 282–299, https://doi.org/10.1175/JAMC-D-13-0128.1.
Chu, X., B. Geerts, L. Xue, and B. Pokharel, 2017a: A case study of cloud radar observations and large eddy simulations of a shallow stratiform orographic cloud, and the impact of glaciogenic seeding. J. Appl. Meteor. Climatol., 56, 1285–1304, https://doi.org/10.1175/JAMC-D-16-0364.1.
Chu, X., B. Geerts, L. Xue, and R. Rasmussen, 2017b: Large eddy simulations of the impact of ground-based glaciogenic seeding on shallow orographic convection: A case study. J. Appl. Meteor. Climatol., 56, 69–84, https://doi.org/10.1175/JAMC-D-16-0191.1.
Crewell, S., and U. Löhnert, 2003: Accuracy of cloud liquid water path from ground-based microwave radiometry 2. Sensor accuracy and synergy. Radio Sci., 38, 8042, https://doi.org/10.1029/2002RS002634.
Daly, C., M. Halbleib, J. I. Smith, W. P. Gibson, M. K. Doggett, G. H. Taylor, J. Curtis, and P. P. Pasteris, 2008: Physiographically sensitive mapping of climatological temperature and precipitation across the conterminous United States. Int. J. Climatol., 28, 2031–2064, https://doi.org/10.1002/joc.1688.
DeMott, P. J., D. C. Rogers, and S. M. Kreidenweis, 1997: The susceptibility of ice formation in upper tropospheric clouds to insoluble aerosol components. J. Geophys. Res., 102, 19 575–19 584, https://doi.org/10.1029/97JD01138.
DeMott, P. J., and Coauthors, 2016: Sea spray aerosol as a unique source of ice nucleating particles. Proc. Natl. Acad. Sci. USA, 113, 5797–5803, https://doi.org/10.1073/pnas.1514034112.
Eilers, P. H. C., and H. F. M. Boelens, 2005: Baseline correction with asymmetric least squares smoothing. Leiden University Medical Centre Rep., 19 pp.
Espitalié, M., 2020: The relationship between the Froude number and the actual flow near real mountains: A case study for the Wind River Range in Wyoming. M.S. thesis, Dept. of Atmospheric Science, University of Wyoming, 62 pp., https://search.proquest.com/pqdtlocal1005743/docview/2453156143/abstract/CE766FC5F9C64C59PQ/1.
Field, P. R., and Coauthors, 2017: Secondary ice production: Current state of the science and recommendations for the future. Ice Formation and Evolution in Clouds and Precipitation: Measurement and Modeling Challenges, Meteor. Monogr., No. 58, Amer. Meteor. Soc., https://doi.org/10.1175/AMSMONOGRAPHS-D-16-0014.1.
Flossmann, A. I., M. Manton, A. Abshaev, R. Bruintjes, M. Murakami, T. Prabhakaran, and Z. Yao, 2019: Review of advances in precipitation enhancement research. Bull. Amer. Meteor. Soc., 100, 1465–1480, https://doi.org/10.1175/BAMS-D-18-0160.1.
French, J. R., and Coauthors, 2018: Precipitation formation from orographic cloud seeding. Proc. Natl. Acad. Sci. USA, 115, 1168–1173, https://doi.org/10.1073/pnas.1716995115.
Friedrich, K., and Coauthors, 2020: Quantifying snowfall from orographic cloud seeding. Proc. Natl. Acad. Sci. USA, 117, 5190–5195, https://doi.org/10.1073/pnas.1917204117.
Geerts, B., and Coauthors, 2013: The AgI Seeding Cloud Impact Investigation (ASCII) campaign 2012: Overview and preliminary results. J. Wea. Modif., 45, 24–43.
Geresdi, I., L. Xue, and R. Rasmussen, 2017: Evaluation of orographic cloud seeding using a bin microphysics scheme: Two-dimensional approach. J. Appl. Meteor. Climatol., 56, 1443–1462, https://doi.org/10.1175/JAMC-D-16-0045.1.
Hausfather, Z., and G. P. Peters, 2020: Emissions—The ‘business as usual’ story is misleading. Nature, 577, 618–620, https://doi.org/10.1038/d41586-020-00177-3.
Hong, S.-Y., Y. Noh, and J. Dudhia, 2006: A new vertical diffusion package with an explicit treatment of entrainment processes. Mon. Wea. Rev., 134, 2318–2341, https://doi.org/10.1175/MWR3199.1.
Ikeda, K., and Coauthors, 2010: Simulation of seasonal snowfall over Colorado. Atmos. Res., 97, 462–477, https://doi.org/10.1016/j.atmosres.2010.04.010.
Jing, X., B. Geerts, Y. Wang, and C. Liu, 2017: Evaluating seasonal orographic precipitation in the interior western United States using gauge data, gridded precipitation estimates, and a regional climate simulation. J. Hydrometeor., 18, 2541–2558, https://doi.org/10.1175/JHM-D-17-0056.1.
Jing, X., B. Geerts, Y. Wang, and C. Liu, 2019: Ambient factors controlling the wintertime precipitation distribution across mountain ranges in the interior western United States. Part II: Changes in orographic precipitation distribution in a pseudo–global warming simulation. J. Appl. Meteor. Climatol., 58, 695–715, https://doi.org/10.1175/JAMC-D-18-0173.1.
Leinonen, J., M. D. Lebsock, G. L. Stephens, and K. Suzuki, 2016: Improved retrieval of cloud liquid water from CloudSat and MODIS. J. Appl. Meteor. Climatol., 55, 1831–1844, https://doi.org/10.1175/JAMC-D-16-0077.1.
Lenaerts, J. T. M., K. Van Tricht, S. Lhermitte, and T. S. L’Ecuyer, 2017: Polar clouds and radiation in satellite observations, reanalyses, and climate models. Geophys. Res. Lett., 44, 3355–3364, https://doi.org/10.1002/2016GL072242.
Li, Z., and R. L. Pitter, 1997: Numerical comparison of two ice crystal formation mechanisms on snowfall enhancement from ground-based aerosol generators. J. Appl. Meteor., 36, 70–85, https://doi.org/10.1175/1520-0450(1997)036<0070:NCOTIC>2.0.CO;2.
Liu, C., K. Ikeda, G. Thompson, R. Rasmussen, and J. Dudhia, 2011: High-resolution simulations of wintertime precipitation in the Colorado headwaters region: Sensitivity to physics parameterizations. Mon. Wea. Rev., 139, 3533–3553, https://doi.org/10.1175/MWR-D-11-00009.1.
Liu, C., and Coauthors, 2017: Continental-scale convection-permitting modeling of the current and future climate of North America. Climate Dyn., 49, 71–95, https://doi.org/10.1007/s00382-016-3327-9.
Lundquist, J., M. Hughes, E. Gutmann, and S. Kapnick, 2019: Our skill in modeling mountain rain and snow is bypassing the skill of our observational networks. Bull. Amer. Meteor. Soc., 100, 2473–2490, https://doi.org/10.1175/BAMS-D-19-0001.1.
Markowski, P., and Y. Richardson, 2011: Mesoscale Meteorology in Midlatitudes. John Wiley and Sons, 658 pp.
Merino, A., and Coauthors, 2019: Aircraft icing: In-cloud measurements and sensitivity to physical parameterizations. Geophys. Res. Lett., 46, 11 559–11 567, https://doi.org/10.1029/2019GL084424.
Minder, J. R., T. W. Letcher, and S. M. Skiles, 2016: An evaluation of high-resolution regional climate model simulations of snow cover and albedo over the Rocky Mountains, with implications for the simulated snow-albedo feedback. J. Geophys. Res., 121, 9069–9088, https://doi.org/10.1002/2016JD024995.
Morrison, A. E., S. T. Siems, and M. J. Manton, 2013: On a natural environment for glaciogenic cloud seeding. J. Appl. Meteor. Climatol., 52, 1097–1104, https://doi.org/10.1175/JAMC-D-12-0108.1.
Muhlbauer, A., and U. Lohmann, 2009: Sensitivity studies of aerosol–cloud interactions in mixed-phase orographic precipitation. J. Atmos. Sci., 66, 2517–2538, https://doi.org/10.1175/2009JAS3001.1.
Pokharel, B., and B. Geerts, 2016: A multi-sensor study of the impact of ground-based glaciogenic seeding on clouds and precipitation over mountains in Wyoming. Part I: Project description. Atmos. Res., 182, 269–281, https://doi.org/10.1016/j.atmosres.2016.08.008.
Pokharel, B., B. Geerts, X. Jing, K. Friedrich, K. Ikeda, and R. Rasmussen, 2017a: A multi-sensor study of the impact of ground-based glaciogenic seeding on clouds and precipitation over mountains in Wyoming. Part II: Seeding impact analysis. Atmos. Res., 183, 42–57, https://doi.org/10.1016/j.atmosres.2016.08.018.
Pokharel, B., B. Geerts, X. Chu, and P. Bergmaier, 2017b: Profiling radar observations and numerical simulations of a downslope wind storm and rotor on the lee of the Medicine Bow Mountains in Wyoming. Atmosphere, 8, 39, https://doi.org/10.3390/atmos8020039.
Pokharel, B., B. Geerts, and X. Chu, 2018: The impact of ground-based glaciogenic seeding on a shallow stratiform cloud over the Sierra Madre in Wyoming: A multi-sensor study of the 3 March 2012 case. Atmos. Res., 214, 74–90, https://doi.org/10.1016/j.atmosres.2018.07.013.
Pokharel, B., and Coauthors, 2021: A modeling examination of cloud seeding conditions under the warmer climate in Utah, USA. Atmos. Res., 248, 105239, https://doi.org/10.1016/j.atmosres.2020.105239.
Rasmussen, R., and Coauthors, 2011: High-resolution coupled climate runoff simulations of seasonal snowfall over Colorado: A process study of current and warmer climate. J. Climate, 24, 3015–3048, https://doi.org/10.1175/2010JCLI3985.1.
Rasmussen, R., and Coauthors, 2014: Climate change impacts on the water balance of the Colorado headwaters: High-resolution regional climate model simulations. J. Hydrometeor., 15, 1091–1116, https://doi.org/10.1175/JHM-D-13-0118.1.
Rasmussen, R., and Coauthors, 2018: Evaluation of the Wyoming Weather Modification Pilot Project (WWMPP) using two approaches: Traditional statistics and ensemble modeling. J. Appl. Meteor. Climatol., 57, 2639–2660, https://doi.org/10.1175/JAMC-D-17-0335.1.
Rauber, R. M., and Coauthors, 2019: Wintertime orographic cloud seeding—A review. J. Appl. Meteor. Climatol., 58, 2117–2140, https://doi.org/10.1175/JAMC-D-18-0341.1.
Saha, S., and Coauthors, 2010: NCEP Climate Forecast System Reanalysis (CFSR) 6-hourly products, January 1979 to December 2010. Research Data Archive at the National Center for Atmospheric Research, Computational and Information Systems Laboratory, accessed 11 August 2014, https://doi.org/10.5065/D69K487J.
Serreze, M. C., M. P. Clark, R. L. Armstrong, D. A. McGinnis, and R. S. Pulwarty, 1999: Characteristics of the western United States snowpack from snowpack telemetry (SNOTEL) data. Water Resour. Res., 35, 2145–2160, https://doi.org/10.1029/1999WR900090.
Solomon, A., G. de Boer, J. M. Creamean, A. McComiskey, M. D. Shupe, M. Maahn, and C. Cox, 2018: The relative impact of cloud condensation nuclei and ice nucleating particle concentrations on phase partitioning in Arctic mixed-phase stratocumulus clouds. Atmos. Chem. Phys., 18, 17 047–17 059, https://doi.org/10.5194/acp-18-17047-2018.
Tan, I., T. Storelvmo, and M. D. Zelinka, 2016: Observational constraints on mixed-phase clouds imply higher climate sensitivity. Science, 352, 224–227, https://doi.org/10.1126/science.aad5300.
Tessendorf, S. A., and Coauthors, 2019: A transformational approach to winter orographic weather modification research: The SNOWIE Project. Bull. Amer. Meteor. Soc., 100, 71–92, https://doi.org/10.1175/BAMS-D-17-0152.1.
Tessendorf, S. A., K. Ikeda, C. Weeks, R. Rasmussen, J. Wolff, and L. Xue, 2020: An assessment of winter orographic precipitation and cloud-seeding potential in Wyoming. J. Appl. Meteor. Climatol., 59, 1217–1238, https://doi.org/10.1175/JAMC-D-19-0219.1.
Thompson, G., P. R. Field, R. M. Rasmussen, and W. D. Hall, 2008: Explicit forecasts of winter precipitation using an improved bulk microphysics scheme. Part II: Implementation of a new snow parameterization. Mon. Wea. Rev., 136, 5095–5115, https://doi.org/10.1175/2008MWR2387.1.
Thompson, G., M. K. Politovich, and R. M. Rasmussen, 2017: A numerical weather model’s ability to predict characteristics of aircraft icing environments. Wea. Forecasting, 32, 207–221, https://doi.org/10.1175/WAF-D-16-0125.1.
Turner, D. D., 2007: Improved ground-based liquid water path retrievals using a combined infrared and microwave approach. J. Geophys. Res., 112, D15204, https://doi.org/10.1029/2007JD008530.
Wang, Y., B. Geerts, and C. Liu, 2018: A 30-year convection-permitting regional climate simulation over the interior western United States. Part I: Validation. Int. J. Climatol., 38, 3684–3704, https://doi.org/10.1002/joc.5527.
Westwater, E. R., Y. Han, M. D. Shupe, and S. Y. Matrosov, 2001: Analysis of integrated cloud liquid and precipitable water vapor retrievals from microwave radiometers during the Surface Heat Budget of the Arctic Ocean project. J. Geophys. Res., 106, 32 019–32 030, https://doi.org/10.1029/2000JD000055.
Xu, M., G. Thompson, D. R. Adriaansen, and S. D. Landolt, 2019: On the value of time-lag-ensemble averaging to improve numerical model predictions of aircraft icing conditions. Wea. Forecasting, 34, 507–519, https://doi.org/10.1175/WAF-D-18-0087.1.
Xue, L., and Coauthors, 2013a: Implementation of a silver iodide cloud-seeding parameterization in WRF. Part I: Model description and idealized 2D sensitivity tests. J. Appl. Meteor. Climatol., 52, 1433–1457, https://doi.org/10.1175/JAMC-D-12-0148.1.
Xue, L., S. A. Tessendorf, E. Nelson, R. Rasmussen, D. Breed, S. Parkinson, P. Holbrook, and D. Blestrud, 2013b: Implementation of a silver iodide cloud-seeding parameterization in WRF. Part II: 3D Simulations of actual seeding events and sensitivity tests. J. Appl. Meteor. Climatol., 52, 1458–1476, https://doi.org/10.1175/JAMC-D-12-0149.1.
Xue, L., and Coauthors, 2020: How will rainfall change over Hawai‘i in the future? High-resolution regional climate simulation of the Hawaiian Islands. Bull. Atmos. Sci. Technol., 1, 459–490, https://doi.org/10.1007/s42865-020-00022-5.
Yau, M. K., and R. R. Rogers, 1996: A Short Course in Cloud Physics. Elsevier, 304 pp.