Numerical evaluation of WRF microphysics parameterizations for a major rainfall event over an arid region

This study examines how cloud microphysical parameterizations in the Weather Research and Forecasting (WRF) model influence the simulation of rainfall over Kuwait during November 2018, with particular focus on the extreme event of 14 November. The WRF model was configured at 4-km resolution and dynamically downscaled from the Community Climate System Model version 4 (CCSM4) to evaluate the performance of four bulk microphysics schemes: WSM6, Lin, Thompson, and Morrison. Model output was evaluated against observations from the Kuwait Automatic Weather Station (29.22°N, 47.96°E) using cumulative and distributional rainfall characteristics, event-scale analysis, and standard statistical metrics (RMSE, MAE, bias, and correlation). All schemes reproduced the timing of rainfall events, while notable differences were found in simulated rainfall intensity. WSM6 produced rainfall amounts closest to observations, with the lowest error values and a small positive bias. Lin showed moderate overestimation, Thompson produced larger overestimation during heavy-rain periods, and Morrison consistently underestimated rainfall totals. For the 14 November event, which recorded 79.0 mm at the station, simulated totals ranged from 55.6 to 91.6 mm across the schemes, indicating that inter-scheme differences were dominated by rainfall magnitude rather than timing. These results highlight the sensitivity of convection-permitting rainfall simulations in arid regions to microphysical formulation. Evaluation using additional events and seasons is needed to assess whether these results extend beyond the period examined.

1 Introduction

Cool-season precipitation over Kuwait and the northeastern Arabian Peninsula is primarily produced by mixed-phase convective systems embedded within large-scale frontal disturbances. These storms are characterized by strong vertical coupling between ice-phase growth aloft and melting-layer processes during descent, followed by substantial evaporation within a deep, dry subcloud layer. As a result, surface rainfall is not solely determined by synoptic forcing, but also by the efficiency of microphysical transformation processes governing riming, aggregation, melting, and raindrop evaporation. The coexistence of vigorous ice-phase production and strong evaporative loss makes rainfall generation in this arid environment particularly sensitive to cloud-microphysical representation in numerical models (Al-Sarmi and Washington, 2011; Almazroui et al., 2017).

During the cool season, rainfall over Kuwait is dynamically controlled by eastward-propagating midlatitude troughs extending from the Eastern Mediterranean. As these troughs deepen, they interact with low-level moisture transported from the Red Sea and Arabian Sea, enhancing frontal ascent and low-level convergence over the northern Arabian Peninsula (Krichak et al., 2016). The resulting vertical motion promotes deep cloud development and embedded mesoscale convective systems, which are responsible for most high-impact rainfall events affecting Kuwait. In November, this dynamical configuration is particularly active as the regional circulation transitions from late autumn to early winter, allowing Mediterranean troughs to penetrate southeastward and interact more frequently with Red Sea Trough extensions, thereby strengthening frontal lifting and organizing convection (Krichak et al., 2016).

In arid environments such as Kuwait, rainfall production is governed by a fundamental competition between ice-phase growth and evaporative loss beneath cloud base. Precipitation embryos generated through riming and aggregation aloft must survive melting and subsequent descent through a deep unsaturated boundary layer, where evaporation can remove a substantial fraction of the falling hydrometeor mass. Consequently, relatively small differences in graupel size, fall speed, melting rate, and raindrop evaporation efficiency can lead to large differences in surface rainfall. These physical controls link precipitation intensity directly to microphysical formulation rather than to synoptic forcing alone and explain why arid-region rainfall simulations often exhibit pronounced sensitivity to the choice of cloud-microphysics parameterization (Morrison et al., 2009; Köcher et al., 2023).

Numerical modeling studies have shown that microphysics schemes exert direct control over storm intensity, hydrometeor partitioning, cold-pool development, and rainfall efficiency in convection-permitting simulations. In mixed-phase systems, differences in riming efficiency, snow–graupel conversion, sedimentation velocity, and rain evaporation modify both the vertical latent-heating structure and the amount of precipitation reaching the surface. Sensitivity experiments conducted over desert and semi-arid environments indicate that single-moment and double-moment schemes can produce substantially different rainfall responses under identical large-scale forcing (Thompson et al., 2008; Morrison et al., 2009; Köcher et al., 2023).

Despite the recognized importance of microphysical processes, sensitivity studies over Kuwait and the northeastern Arabian Peninsula remain limited. Most previous numerical investigations in the region have focused on synoptic-scale circulation, moisture transport pathways, and thermodynamic instability associated with Mediterranean troughs, Red Sea Trough dynamics, and subtropical forcing (Al-Sarmi and Washington, 2011; Almazroui et al., 2017). In contrast, scheme-dependent precipitation physics—particularly the role of ice-phase growth, melting efficiency, and subcloud evaporation—has received comparatively limited attention. As a result, the influence of microphysics selection on event-scale rainfall magnitude over Kuwait is not yet well documented in the modeling literature (Taraphdar et al., 2022).

The November 2018 rainfall episode represents one of the most intense cool-season precipitation events observed over Kuwait in recent decades. A sequence of frontal disturbances produced widespread rainfall across the country, culminating on 14 November when 79.0 mm was recorded within 24 h at the Kuwait Automatic Weather Station. This event accounted for most of the monthly accumulation and provides a representative example of dynamically forced, mixed-phase convection operating within a strongly evaporative desert boundary layer.

Sensitivity experiments are conducted using the WRF model (Skamarock et al., 2019) with dynamic downscaling from the CCSM4 (Gent et al., 2011). Four bulk microphysics parameterizations are evaluated: the Lin scheme (Lin et al., 1983), the WSM6 scheme (Hong and Lim, 2006), the Thompson scheme (Thompson et al., 2008), and the Morrison double-moment scheme (Morrison et al., 2009). These WRF schemes span traditional single-moment formulations to prognostic number-concentration approaches and differ in their treatment of ice-phase growth, riming processes, melting characteristics, and rain evaporation. Because these formulations directly control hydrometeor size distributions and sedimentation behavior, they are expected to produce different rainfall responses under identical dynamical forcing.

This study examines how microphysics scheme formulation influences cumulative rainfall, peak intensity, and event-scale precipitation characteristics during November 2018. Particular attention is given to whether inter-scheme differences arise primarily from rainfall magnitude rather than timing errors and to relating these differences to underlying microphysical process representations in a dry, mixed-phase convective environment.

2 Methods

2.1 WRF configuration

Four numerical simulations were conducted using the Weather Research and Forecasting (WRF) model version 4.5 (Skamarock et al., 2019). A single domain (Figure 1) was configured over Kuwait with a horizontal grid spacing of 4 km and 50 vertical eta levels extending to 50 hPa. The model domain was centered over Kuwait and extended over the surrounding region relevant to the synoptic-scale circulation influencing the country, including the northern Arabian Peninsula and adjacent portions of the Arabian Gulf.

Figure 1

Figure 1. WRF model domain at 4 km horizontal resolution centered over Kuwait.

The simulations covered the period 01–30 November 2018, initialized at 0000 UTC on 01 November. The first 24 h were excluded from the analysis as model spin-up. The month-long integration allowed representation of the evolving large-scale circulation and thermodynamic conditions associated with the November 2018 rainfall episode, including the extreme precipitation observed on 14 November.

Initial and lateral boundary conditions were obtained from the Community Climate System Model version 4 (CCSM4) reforecast dataset (Gent et al., 2011), available at 1° horizontal resolution with 6-hourly updates. The simulations were conducted using three nested domains with horizontal grid spacings of 36, 12, and 4 km, consistent with a 3:1 nesting ratio from the 1° IC/BC resolution. This corresponds to an IC/BC-to-domain resolution ratio of approximately 28:1, consistent with common practice in regional dynamical downscaling. CCSM4 atmospheric fields are provided on 26 hybrid sigma-pressure levels and were interpolated to the model’s 50 eta levels using the WRF Preprocessing System (WPS). Interpolated variables included temperature, geopotential height, horizontal winds, specific humidity, and surface pressure. Sea-surface temperatures and land/soil fields were also taken from CCSM4 to ensure consistency between lower-boundary forcing and atmospheric fields.

Initial and lateral boundary conditions were derived from CCSM4, which represents the large-scale circulation associated with cool-season precipitation over the northern Arabian Peninsula. Previous regional downscaling studies over Kuwait have demonstrated that CCSM4 reproduces realistic thermodynamic structures and precipitation characteristics, supporting its use as forcing for the present simulations (Alsarraf, 2022, 2024).

At a grid spacing of 4 km, deep convection is represented explicitly by the model dynamics. Therefore, the cumulus parameterization was disabled in all simulations to ensure that precipitation developed solely through resolved cloud and microphysical processes. Previous studies have shown that grid spacings of approximately 3–4 km are sufficient to explicitly represent deep convection over arid and semi-arid regions without the use of a cumulus parameterization (Taraphdar et al., 2022; Köcher et al., 2023).

WRF model output was evaluated using daily rainfall observations from the Automatic Weather Station (AWOS) located at 29.22°N, 47.96°E in Kuwait. The station provides a continuous record for November 2018 and captured the maximum rainfall associated with the 14 November extreme event. Observed precipitation was aggregated to 24-h totals to match the temporal resolution of the WRF diagnostic output.

2.2 Physics schemes

Four microphysical schemes were selected to evaluate their response and identify the best-performing scheme for convection-permitting WRF simulations over an arid region. Precipitation in arid environments is strongly influenced by microphysical processes, particularly hydrometeor size distributions, evaporation below cloud base, and mixed-phase processes aloft, making the choice of microphysics a critical source of uncertainty.

The Lin and WSM6 schemes are single-moment mixed-phase parameterizations that include graupel and predict only hydrometeor mass. These schemes impose fixed assumptions on particle number and size, which can influence precipitation efficiency and evaporation in dry sub-cloud layers and provide a useful baseline for assessing the impact of simplified microphysical representations in arid convection. The Thompson scheme represents an intermediate level of complexity, incorporating variable intercept parameters and a more flexible treatment of mixed-phase processes, which can affect hydrometeor growth and fallout in environments characterized by dry boundary layers. The Morrison scheme is a double-moment parameterization that predicts both mass and number concentrations for multiple hydrometeor species, allowing explicit variation in particle size distributions and enhanced sensitivity to evaporation and sedimentation processes that are expected to be important in arid convective environments.

These schemes were selected to span increasing levels of microphysical complexity and to isolate the influence of key microphysical assumptions on simulated precipitation in an arid environment, while remaining computationally feasible for month-long simulations under identical large-scale forcing.

All WRF physical parameterizations were kept identical across simulations to isolate the effects of microphysics. The planetary boundary layer was represented by the Yonsei University (YSU) scheme (Hong et al., 2006). Surface fluxes were computed using Monin–Obukhov similarity theory. Land-surface processes were simulated using the Noah land surface model with four soil layers (Tewari et al., 2004). Longwave radiation was parameterized using the RRTMG scheme, while shortwave radiation was calculated using the Dudhia scheme (Iacono et al., 2008) (Table 1).

Table 1

Table 1. WRF physics and parameterization schemes.

3 Results

3.1 WRF rainfall results

Monthly rainfall during November 2018 is characterized by several distinct precipitation episodes, including an early-month event between 6 and 10 November, a pronounced mid-month peak, and a weaker event near the end of the month (Figure 2). All WRF simulations reproduce the timing of these events, indicating consistent representation of large-scale forcing across microphysical schemes. Differences among simulations are primarily associated with rainfall magnitude rather than event timing.

Figure 2

Figure 2. Daily rainfall at Kuwait AWOS for November 2018 shown as grouped columns for observed, WSM6, Lin, Morrison, and Thompson.

The Thompson scheme produces the highest rainfall amounts throughout the month, with particularly strong overestimation during the mid-month event visible in the daily time series (Figure 2). Lin also overestimates rainfall, though to a lesser extent. In contrast, Morrison consistently produces lower rainfall totals and underestimates peak intensities, while WSM6 shows the closest agreement with observed daily rainfall, exhibiting only a modest positive bias. Dry periods are well simulated by all schemes.

These differences accumulate over the month, leading to divergence in cumulative rainfall after the mid-month event (Figure 3). Observed monthly rainfall at Kuwait AWOS is 261.8 mm, while simulated totals range from 222.4 mm for Morrison to 323.9 mm for Thompson, with intermediate values of 273.3 mm for WSM6 and 291.2 mm for Lin. The cumulative curves maintain a consistent ranking through the remainder of the month, with Thompson producing the largest accumulation and Morrison the smallest.

Figure 3

Figure 3. Cumulative rainfall at Kuwait AWOS during November 2018 from observations and WRF simulations using the WSM6, Lin, Morrison, and Thompson microphysics schemes.

Daily rainfall variability relative to observations further reflects these systematic differences (Figure 4). WSM6 and Lin generally cluster near the 1:1 line across a wide range of rainfall amounts, whereas Thompson departs toward higher values during intense rainfall and Morrison frequently underestimates daily totals. Although median daily rainfall values are similar across schemes due to the prevalence of dry days, the spread differs substantially, with Thompson exhibiting the widest range and Morrison the narrowest.

Figure 4

Figure 4. Observed AWOS versus WRF daily rainfall at Kuwait for November 2018 for the WSM6, Lin, Morrison, and Thompson schemes.

Pattern statistics summarized by the Taylor diagram (Figure 5) show high correlation between simulations and observations for all schemes, while differences are mainly associated with variance. WSM6 and Lin lie closest to the observational reference, Thompson exhibits enhanced variance, and Morrison shows reduced variance. These characteristics are consistent with the error metrics summarized in Table 2 and Figure 6, which indicate the lowest RMSE and MAE for WSM6, followed by Lin, and substantially larger errors for Thompson and Morrison.

Figure 5

Figure 5. Taylor diagram comparing daily rainfall in November 2018 at Kuwait AWOS for the WSM6, Lin, Morrison, and Thompson schemes.

Table 2

Table 2. Error metrics for daily rainfall at Kuwait AWOS, November 2018—RMSE, MAE, bias and correlation for WSM6, Lin, Morrison and Thompson; RMSE, MAE and bias in mm.

Figure 6

Figure 6. Monthly error metrics for daily rainfall at Kuwait AWOS during November 2018 for the WSM6, Lin, Morrison, and Thompson schemes.

3.2 Extreme rainfall event of 14 November

The most intense rainfall during the study period occurred on 14 November 2018, when observed daily accumulation reached 79.0 mm at Kuwait AWOS. All WRF simulations reproduce the occurrence of this event on the correct day, indicating consistent representation of the synoptic forcing. However, substantial differences are evident in the simulated rainfall magnitude among the microphysical schemes. WSM6 produces an accumulated rainfall of 81.2 mm, in close agreement with observations, while Lin slightly overestimates the event with 84.3 mm. Thompson generates the highest rainfall amount (91.6 mm), whereas Morrison substantially underestimates the event, producing only 55.6 mm. These differences are reflected in the relative errors, which range from +2.8% for WSM6 and +6.7% for Lin to +15.9% for Thompson and −29.6% for Morrison. Residuals expressed in physical units follow the same ranking, with the smallest errors associated with WSM6 and the largest underestimation associated with Morrison. Overall, the 14 November event highlights strong sensitivity of simulated extreme rainfall intensity to the choice of microphysical scheme, despite consistent event timing across all simulations (see Figures 7–9).

Figure 7

Figure 7. Rainfall on 14 November 2018 at Kuwait AWOS compared with WRF simulations using WSM6, Lin, Morrison, and Thompson.

Figure 8

Figure 8. Percent error relative to the observed 79.0 mm on 14 November 2018 at Kuwait AWOS for WSM6, Lin, Morrison, and Thompson.

Figure 9

Figure 9. Residuals (model − observed) on 14 November 2018 at Kuwait AWOS for WSM6, Lin, Morrison, and Thompson (mm).

4 Discussion

Rainfall intensity over arid regions exhibits pronounced sensitivity to cloud microphysical parameterizations, even when event timing and large-scale forcing are consistently represented. In the November 2018 simulations over Kuwait, all WRF experiments reproduced the occurrence and timing of precipitation episodes (Figure 2), indicating that synoptic-scale forcing was well captured across configurations. Differences among simulations therefore arise primarily from microphysical controls on precipitation efficiency rather than from deficiencies in large-scale dynamics.

The thermodynamic environment during the study period is characteristic of the Arabian Peninsula, with a deep melting layer and a strongly unsaturated boundary layer. Under such conditions, surface rainfall depends critically on hydrometeor size, fall speed, and evaporation during descent. Previous studies have shown that variations in these processes can produce large differences in simulated precipitation, particularly in dry environments where subcloud evaporation is strong (Gilmore et al., 2004; Milbrandt and Yau, 2005).

The divergence in rainfall magnitude among the microphysics schemes (Figures 2–4) reflects differences in how each parameterization represents mixed-phase growth, sedimentation, and evaporation. Schemes that promote efficient riming and produce larger precipitation particles experience reduced evaporative losses, leading to higher surface rainfall totals. Conversely, schemes that generate smaller hydrometeors or weaker sedimentation are more susceptible to evaporation within deep dry subcloud layers, resulting in systematic underestimation of rainfall. This behavior is consistent with prior idealized and real-case studies demonstrating that hydrometeor size distributions and fall speeds exert first-order control on precipitation efficiency (Morrison and Milbrandt, 2011; Bao et al., 2019).

Within this framework, the comparatively strong performance of the WSM6 scheme during November 2018 (Figures 3–6) can be attributed to its graupel-based single-moment formulation, fixed intercept parameters, and relatively conservative riming efficiency. These characteristics favor larger effective hydrometeor sizes and faster sedimentation, limiting evaporative losses without producing unrealistically high rainfall intensities. As a result, WSM6 yields smaller overall biases at both the monthly scale and during the extreme event of 14 November (Figure 6). Similar advantages of graupel-based single-moment schemes have been reported in dry and semi-arid environments, where evaporation below cloud base strongly modulates surface rainfall (Gilmore et al., 2004; Van Weverberg, 2013).

In contrast, the Thompson scheme produces higher rainfall totals, particularly during intense events (Figures 2, 4, 6), reflecting stronger mixed-phase growth and enhanced precipitation efficiency. While this behavior can improve rainfall representation in moist environments, it leads to overestimation under arid boundary-layer conditions where evaporation is significant. The Morrison scheme consistently underestimates rainfall in both monthly accumulation and extreme-event intensity (Figures 3, 6), likely due to smaller hydrometeor sizes and reduced sedimentation efficiency associated with its two-moment formulation. Similar underestimation has been documented in previous evaluations of two-moment schemes under dry thermodynamic conditions (Morrison and Milbrandt, 2011; Bao et al., 2019).

The analysis further demonstrates that these sensitivities are not confined to a single event. Although the study period spans only 1 month, the consistent ranking of scheme performance across multiple precipitation episodes (Figures 2–4) indicates that the identified differences are systematic and linked to the interaction between microphysical assumptions and the arid environment. This finding aligns with prior work showing that environmental humidity modulates the relative performance of microphysics schemes and that no single parameterization is universally optimal across climates (Van Weverberg, 2013).

Overall, the results emphasize that accurate simulation of rainfall over arid regions requires microphysics schemes that realistically represent hydrometeor growth and evaporation within deep dry boundary layers. Selection of microphysics schemes without consideration of environmental thermodynamics can lead to systematic biases in precipitation magnitude, even when storm timing and synoptic evolution are well simulated.

5 Conclusion

This study evaluated the performance of four WRF microphysics schemes for simulating rainfall over Kuwait during November 2018, including the extreme event of 14 November. All schemes reproduced the timing of rainfall events, while notable differences were found in simulated rainfall magnitude. At both the monthly scale and during the extreme event, WSM6 produced rainfall amounts closest to observations. Lin showed moderate positive bias, Thompson produced larger overestimation during high-intensity periods, and Morrison consistently underestimated rainfall totals.

The differences among schemes are linked to how each parameterization represents mixed-phase processes, hydrometeor size distributions, and evaporation within the deep dry boundary layer typical of arid environments. These processes exert strong control on surface rainfall amounts, leading to systematic differences in simulated precipitation despite similar large-scale forcing and storm evolution. For the conditions examined here, the WSM6 formulation yielded smaller overall biases relative to the other schemes.

The conclusions of this study are based on a single month containing multiple precipitation events and should not be interpreted as a general ranking of microphysical schemes. Further evaluation using additional events, different synoptic conditions, and expanded observational datasets, including radar observations, is needed to assess whether the relative performance identified here persists across other rainfall episodes in Kuwait and the broader Arabian Peninsula.

Data availability statement

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

Author contributions

HA: Software, Visualization, Data curation, Methodology, Resources, Formal analysis, Investigation, Supervision, Writing – review & editing, Project administration, Validation, Funding acquisition, Writing – original draft, Conceptualization.

Funding

The author(s) declared that financial support was not received for this work and/or its publication.

Conflict of interest

The author(s) declared that this work was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Generative AI statement

The author(s) declared that Generative AI was not used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

References

Almazroui, M., Islam, M. N., Saeed, S., Alkhalaf, A. K., and Dambul, R. (2017). Assessment of uncertainties in projected temperature and precipitation over the Arabian peninsula using three categories of CMIP5 multimodel ensembles. Earth Syst. Environ. 1, 1–20. doi: 10.1007/s41748-017-0027-5

Crossref Full Text | Google Scholar

Al-Sarmi, S., and Washington, R. (2011). Recent observed climate change over the Arabian peninsula. J. Geophys. Res. Atmos. 116:D11109. doi: 10.1029/2010JD015459

Crossref Full Text | Google Scholar

Alsarraf, H. (2022). Projected climate change over Kuwait simulated using a WRF high-resolution regional climate model. Int. J. Glob. Warm. 26, 198–211. doi: 10.1504/ijgw.2022.120844

Crossref Full Text | Google Scholar

Alsarraf, H. (2024). Dynamic downscaling of climate projections using WRF–CCSM4: future soil temperature predictions for the Arabian peninsula and Kuwait. Front. Clim. 6:1511608. doi: 10.3389/fclim.2024.1511608

Crossref Full Text | Google Scholar

Bao, J.-W., Michelson, S. A., and Grell, E. D. (2019). Microphysical process comparison of three bulk microphysics schemes in the WRF model for an idealized squall-line case. Mon. Weather Rev. 147, 3093–3120. doi: 10.1175/MWR-D-18-0249.1

Crossref Full Text | Google Scholar

Gent, P. R., Danabasoglu, G., Donner, L. J., Holland, M. M., Hunke, E. C., Jayne, S. R., et al. (2011). The community climate system model version 4. J. Clim. 24, 4973–4991. doi: 10.1175/2011JCLI4083.1

Crossref Full Text | Google Scholar

Gilmore, M. S., Straka, J. M., and Rasmussen, E. N. (2004). Precipitation uncertainty due to variations in precipitation particle parameters within a simple microphysics scheme. Mon. Weather Rev. 132, 2610–2627. doi: 10.1175/MWR2810.1

Crossref Full Text | Google Scholar

Hong, S.-Y., Noh, Y., and Dudhia, J. (2006) A new vertical diffusion package with an explicit treatment of entrainment processes. Monthly Weather Review, 134, 2318–2341. doi: 10.1175/MWR3199.1

Crossref Full Text | Google Scholar

Hong, S.-Y., and Lim, J.-O. J. (2006). The WRF single-moment 6-class microphysics scheme (WSM6). J. Korean Meteorol. Soc. 42, 129–151.

Google Scholar

Iacono, M. J., Delamere, J. S., Mlawer, E. J., Shephard, M. W., Clough, S. A., and Collins, W. D. (2008). Radiative forcing by long-lived greenhouse gases: Calculations with the AER radiative transfer models. Journal of Geophysical Research: Atmospheres, 113:D13103. doi: 10.1029/2008JD009944

Crossref Full Text | Google Scholar

Köcher, G., Zinner, T., and Knote, C. (2023). Influence of cloud microphysics schemes on weather model predictions of heavy precipitation. Atmos. Chem. Phys. 23, 6255–6269. doi: 10.5194/acp-23-6255-2023

Crossref Full Text | Google Scholar

Krichak, S. O., Feldstein, S. B., Alpert, P., Gualdi, S., Scoccimarro, E., and Yano, J.-I. (2016). Discussing the role of tropical and subtropical moisture sources in cold-season extreme precipitation events in the Mediterranean region from a climate change perspective. Nat. Hazards Earth Syst. Sci. 16, 269–285. doi: 10.5194/nhess-16-269-2016

Crossref Full Text | Google Scholar

Lin, Y.-L., Farley, R. D., and Orville, H. D. (1983). Bulk parameterization of the snow field in a cloud model. J. Clim. Appl. Meteorol. 22, 1065–1092. doi: 10.1175/1520-0450(1983)022<1065:BPOTSF>2.0.CO;2

Crossref Full Text | Google Scholar

Milbrandt, J. A., and Yau, M. K. (2005). A multimoment bulk microphysics parameterization. Part I: analysis of the role of the spectral shape parameter. J. Atmos. Sci. 62, 3051–3064. doi: 10.1175/JAS3534.1

Crossref Full Text | Google Scholar

Morrison, H., and Milbrandt, J. A. (2011). Comparison of two-moment bulk microphysics schemes in idealized supercell thunderstorm simulations. Mon. Weather Rev. 139, 1103–1130. doi: 10.1175/2010MWR3433.1

Crossref Full Text | Google Scholar

Morrison, H., Thompson, G., and Tatarskii, V. (2009). Impact of cloud microphysics on the development of trailing stratiform precipitation in a simulated squall line: comparison of one- and two-moment schemes. Mon. Weather Rev. 137, 991–1007. doi: 10.1175/2008MWR2556.1

Crossref Full Text | Google Scholar

Skamarock, W. C., Klemp, J. B., Dudhia, J., Gill, D. O., Liu, Z., Berner, J., et al. (2019). A description of the advanced research WRF version 4. NCAR Tech. Note NCAR/TN-556+STR. 1–165. doi: 10.5065/1DFH-6P97

Crossref Full Text | Google Scholar

Taraphdar, S., Pauluis, O. M., Xue, L., Liu, C., Rasmussen, R. M., Ajayamohan, R. S., et al. (2022). WRF gray-zone simulations of precipitation over the Middle East and the UAE: impacts of physical parameterizations and resolution. J. Geophys. Res. Atmos. 127:e2021JD034648. doi: 10.1029/2021JD034648

Crossref Full Text | Google Scholar

Tewari, M., Chen, F., Wang, W., Dudhia, J., LeMone, M. A., Mitchell, K., et al. (2004). Implementation and verification of the unified NOAH land surface model in the WRF model. 20th Conference on Weather Analysis and Forecasting/16th Conference on Numerical Weather Prediction, Seattle, WA: American Meteorological Society, 11–15.

Google Scholar

Thompson, G., Field, P. R., Rasmussen, R. M., and Hall, W. D. (2008). Explicit forecasts of winter precipitation using an improved bulk microphysics scheme. Part II: implementation of a new snow parameterization. Mon. Weather Rev. 136, 5095–5115. doi: 10.1175/2008MWR2387.1

Crossref Full Text | Google Scholar

Van Weverberg, K. (2013). Impact of environmental instability on convective precipitation uncertainty associated with rimed ice species in a bulk microphysics scheme. Mon. Weather Rev. 141, 2841–2849. doi: 10.1175/MWR-D-13-00036.1

Crossref Full Text | Google Scholar