Performance Of Weather Research And Forecasting Research Articles

Fidelity of WRF model in simulating heat wave events over India

The evaluation of Weather Research and Forecasting (WRF) model has been performed for simulating episodic Heat Wave (HW) events of 2015 and 2016 with varied horizontal resolutions of 27 km for the entire India (d01), 9 km for the North West (NW (d02)) and South East (SE (d03)) domain. Study compares the maximum temperature (Tmax) simulated by WRF model, using six different combination of parameterization schemes, with observations from the India Meteorological Department (IMD) during the HW events. Among the six experiments, Exp2 (i.e., combination of WSM6 microphysics (MP) together with radiation parameterization CAM, Yonsei (PBL), NOAH land surface and Grell-3D convective schemes) is found closest to the observations in reproducing the temperature. The model exhibits an uncertainty of ± 2 °C in maximum temperature (Tmax) for both the regions, suggesting regional temperature is influenced by the location and complex orography. Overall, statistical results reveal that the best performance is achieved with Exp2. Further, to understand the dynamics of rising HW intensity, two case studies of HW days along with influencing parameters like Tmax, RH and prevailing wind distribution have been simulated. Model simulated Tmax during 2015 reaches up to 44 °C in NW and SE part of India. In 2016, HW is more prevailing towards NW, while in SE region Tmax reaches upto 34–38 °C with high RH (60–85%). The comparative research made it abundantly evident that these episodic events are unique in terms of duration and geographical spread which can be used to assess the WRF performance for future projections of HW.

Evaluating the performance of WRF in simulating winds and surface meteorology during a Southern California wildfire event

The intensity and frequency of wildfires in California (CA) have increased in recent years, causing significant damage to human health and property. In October 2007, a number of small fire events, collectively referred to as the Witch Creek Fire or Witch Fire started in Southern CA and intensified under strong Santa Ana winds. As a test of current mesoscale modeling capabilities, we use the Weather Research and Forecasting (WRF) model to simulate the 2007 wildfire event in terms of meteorological conditions. The main objectives of the present study are to investigate the impact of horizontal grid resolution and planetary boundary layer (PBL) scheme on the model simulation of meteorological conditions associated with a Mega fire. We evaluate the predictive capability of the WRF model to simulate key meteorological and fire-weather forecast parameters such as wind, moisture, and temperature. Results of this study suggest that more accurate predictions of temperature and wind speed relevant for better prediction of wildfire spread can be achieved by downscaling regional numerical weather prediction products to 1 km resolution. Furthermore, accurate prediction of near-surface conditions depends on the choice of the planetary boundary layer parameterization. The MYNN parameterization yields more accurate prediction as compared to the YSU parameterization. WRF simulations at 1 km resolution result in better predictions of temperature and wind speed than relative humidity during the 2007 Witch Fire. In summary, the MYNN PBL parameterization scheme with finer grid resolution simulations improves the prediction of near-surface meteorological conditions during a wildfire event.

WRF prediction of an atmospheric river‐related precipitation event: Sensitivity to cumulus parameterization schemes

AbstractThe present study aimed to evaluate performance sensitivity to the Cumulus Parameterization Scheme (CPS) used for the Weather Research and Forecasting (WRF) model to predict the atmospheric river‐related precipitation (ARP) event with 206 and 57 mm, highest and area‐averaged precipitation (AAP) per 24‐h, respectively, that occurred over the central mountainous basins of Iran on 31 March 2019. For this purpose, experiments were designed using the 12 (almost all) CPSs available in WRF v4. To verify the predicted precipitation (from the inner 4‐km domain), both point‐scale and grid‐scale comparisons were performed against gauge‐ and satellite‐based observational data at three accumulation time‐scales (12‐, 18‐, and 24‐h) and in three distinct sub‐regions. All scores obtained from the different statistical metrics used, are in complete agreement with a strongly dependent performance of WRF on the CPS used. In addition, the use of Kain‐Fritsch, KF‐CuP, and Grell‐3 CPSs could provide a realistic picture of impending heavy precipitation for WRF. Contrary, the New SAS, Tiedtke, and Zhang‐McFarlane CPSs did not perform satisfactorily in predicting the ARP event. As a result, CPSs with the “momentum transport” option in their modification mechanism are unlikely to adequately simulate the conversion of incoming low‐level moisture from atmospheric river to precipitation. However, precipitation predictions are more accurate at the 24‐h accumulation time‐scale than at the 12‐ and 18‐h. Also, a dry bias in the predictions is expected as the terrain elevation and accumulation time‐scale decrease and the distance from the core of the precipitation field increases.

Performance Evaluation and Optimization of the Weather Research and Forecasting (WRF) Model Based on Kunpeng 920

The Weather Research and Forecasting (WRF) model is a mesoscale numerical weather prediction system, which is widely used in major high-performance server platforms. This study focuses on the performance evaluation and optimization of WRF on Huawei’s self-developed kunpeng 920 processor platform, aiming to improve the operational efficiency of WRF. The results of the study show that the scalability of WRF on kunpeng 920 processor is well performed; the performance of WRF on kunpeng 920 processor is improved by 32.6% after invoking the Fast Math Library and Domain Decomposition Core Tile Division optimization. In terms of IO, the main optimizations are parallel IO and asynchronous IO. Eventually, the single output time of WRF is reduced from 37.28 s in serial IO mode to 0.14 s in asynchronous IO mode, and the overall running time is reduced from 1078.80 s to 807.94 s.

Evaluation of WRF Performance in Simulating an Extreme Precipitation Event over the South of Minas Gerais, Brazil

Extreme precipitation events are becoming increasingly frequent and intense in southeastern Brazil, leading to socio-economic problems. While it is not possible to control these events, providing accurate weather forecasts can help society be better prepared. In this study, we assess the performance of the Weather Research and Forecasting (WRF) model in simulating a period of extreme precipitation from 31 December 2021 to 2 January 2022 in the southern region of Minas Gerais (SMG) state in southeastern Brazil. We conducted five simulations using two nested grids: a 12 km grid (coarse resolution) and a 3 km grid (high resolution). For the coarse resolution, we tested the performance of five cumulus convection parameterization schemes: Kain–Fritsch, Betts–Miller–Janjic, Grell–Freitas, Grell–Devenyi, and New Tiedke. We evaluated the impact of these simulations on driving the high-resolution simulations. To assess the performance of the simulations, we compared them with satellite estimates, in situ precipitation measurements from thirteen meteorological stations, and other variables from ERA5 reanalysis. Based on the results, we found that the Grell–Freitas scheme has better performance in simulating the spatial pattern and intensity of precipitation for the studied region when compared with the other four analyzed schemes.

Performance of the Weather Research and Forecasting model in simulating climate over northern Asia

AbstractHigh‐resolution simulation is important for understanding climate change and its impact over regions with complex terrain. Northern Asia (NA) is such a region and sensitive to global warming. However, the performance of the regional model over NA is still unclear. In this study, we conducted a two‐decade‐long high‐resolution (25 km) simulation over NA using the Weather Research and Forecasting (WRF) model forced by the ERA‐Interim reanalysis and then evaluated the performance of WRF against the observations. The results show that the WRF simulation, set up with selections of the physical schemes, can reasonably reproduce the spatial and temporal features of surface air temperature and precipitation over NA. In addition, the WRF simulation can add more regional details on the climate features than the ERA‐Interim data, especially over the NA regions with complex terrain. The spatial‐pattern correlation coefficient of summer precipitation over NA between the WRF simulation and observations is higher than that of the ERA‐Interim data. WRF can reasonably reproduce the annual cycles of precipitation. Moreover, the observed year‐to‐year variations in summer surface air temperature and precipitation over most NA subregions could be generally captured by the WRF. In conclusion, although systemic model bias still exists, the evaluations in this study indicate that WRF shows reasonable performance in a high‐resolution simulation of climate over NA, consequently providing a scientific basis for future simulation studies over the region.

Time-variant post-processing method for long-term numerical wind speed forecasts based on multi-region recurrent graph network

Weather Research and Forecasting (WRF) is widely used for long-term wind speed prediction. To reduce the inherent systematic error of the WRF, a graph-based wind speed prediction model is proposed by post-processing the WRF. The proposed model contains four stages. In stage 1, the WRF model is applied to generate raw wind speed prediction results. In stage 2, Planar Maximally Filtered Graph (PMFG) is used to construct an informative graph of the wind field. In stage 3, Deep Autoencoder-like Nonnegative Matrix Factorization (DANMF) and Subgraph Alignment and Region Organization (SARO) are utilized to divide the graph into several regions. In stage 4, Multi-region Recurrent Graph Network (MRGN) model is proposed to build multiple regional models, aggregate them via time-variant ensemble weights, and generate improved long-term wind speed prediction results. The wind speed prediction results on 25 real meteorological monitoring nodes show that: (1) the proposed model outperforms WRF and 3 state-of-the-art models with a 95% confidence level; (2) the proposed model tends to produce better performance in strongly dynamic wind speed; (3) the proposed model has enough computational efficiency for practice. In conclusion, the proposed model is effective to improve WRF performance and has application potential.

Comparison of the Forecast Performance of WRF Using Noah and Noah-MP Land Surface Schemes in Central Asia Arid Region

The land surface scheme (LSS) plays a very important role in the forecast of surface meteorological elements in the Central Asia arid region. Therefore, two LSSs viz. Noah and Noah-MP were evaluated over the Central Asia arid region in January and July 2017 using the Weather Research and Forecasting (WRF) model at a 3-km horizontal grid resolution. The objective was to assess the performance of WRF LSSs by calculated the mean error (ME) and the root mean squared error (RMSE) of simulated hourly meteorological elements, such as 2-m air temperature, 10-m wind speed, soil temperature, soil moisture at 5 and 25 cm thickness, and surface soil heat flux at 5 cm thickness. The results showed that, compared to Noah, Noah-MP modeled less surface sensible heat flux in the northern Xinjiang (15~20 W∙m−2) and surface latent heat flux in most areas of Xinjiang (<10 W∙m−2) in January, and mainly generated less sensible heat flux in most areas of north Xinjiang and the mountainous regions of southern Xinjiang (≤20 W·m−2) which on the contrary, generated more surface latent heat flux in most parts of Xinjiang (15~20 W·m−2) in July. Meanwhile, the surface soil heat flux generated from Noah-MP was closer to the observations at Hongliuhe and Kelameili stations in January, the ME increased by 17.5% and reduced by 80.7%, respectively, the RMSE decreased by 44.4% and 61.7%, respectively, and closer to the observations at Xiaotang station in July, the ME and RMSE reduced by 19.1% and 20.5%, respectively. Compared to Noah, Noah-MP improved the overall simulation of soil temperature and soil moisture over the northern and eastern Xinjiang (at 10 cm thickness), the ME and RMSE of simulated soil temperature reduced by 85.0% and 13.4% in January, decreased by 78.6% and increased by 6.2% in July, respectively, and the ME and RMSE of simulated soil moisture reduced by 67.2% and 14.9% in January, reduced by 33.3% and 2.8% in July, respectively. Compared to Noah, Noah-MP’s results were lowered for the simulated 10-m wind speed and 2-m air temperature, especially the simulated 2-m air temperature over the cold climate regions of northern Xinjiang, was improved significantly, the ME and RMSE of simulated 10-m wind speed reduced by 0.8% and 4.9% in January, decreased by 6.7% and 2.8% in July, respectively, the ME and RMSE of simulated 2-m air temperature reduced by 2.8% and 1.0% in July, respectively. This study demonstrated the advantage of coupled Noah-MP over the Central Asia arid region, providing the basis for WRF/Noah-MP in future operational applications in the Central Asia arid region.

WRF hourly evaluation for extreme precipitation events

Precipitation is one of the most relevant fields in atmospheric modeling because of its environmental, social and economic implications. However, precipitation validation from weather model outputs presents substantial challenges, such as measurement uncertainties, use of gridded datasets vs. direct observations, and the selection of statistical goodness-of-fit measures. The main difficulty of working with precipitation is that it can be spatially irregular, especially in extreme events. High temporal aggregation smooths the field and reduces verification uncertainty. For this reason, validations are usually focused on a daily scale. However, many extreme events occur on shorter periods, for which a sub-daily precipitation assessment is required. In this paper, hourly precipitation verification of the Weather Research and Forecasting (WRF) model is explored for 45 extreme precipitation events (EPEs) recorded in northeastern Spain. For this, stations with recorded EPEs were classified according to the hourly distribution of precipitation. WRF simulations were established considering three microphysics and two planetary boundary layer (PBL) parameterizations. Finally, several statistical goodness-of-fit measures and spatial and temporal precipitation distributions were used for evaluating WRF performance. The results showed that microphysics were more important than PBL parameterizations. Goddard and Thompson together with Mellor-Yamada-Nakanishi and Nino PBL gave better results for most of the analyzed characteristics. However, an optimal combination of parameterizations was not obtained for all EPEs, because event characteristics had important effects on model performance.

Ensemble-Tailored Pattern Analysis of High-Resolution Dynamically Downscaled Precipitation Fields: Example for Climate Sensitive Regions of South America

For climate adaptation and risk mitigation, decision makers in water management or agriculture increasingly demand for regionalized weather and climate information. To provide these, regional atmospheric models, such as the Weather Research and Forecasting (WRF) model, need to be optimized in their physical setup to the region of interest. The objective of this study is to evaluate four cumulus physics (CU), two microphysics (MP), two planetary boundary layer physics (PBL), and two radiation physics (RA) schemes in WRF according to their performance in dynamically downscaling the precipitation over two typical South American regions: one orographically complex area in Ecuador/Peru (horizontal resolution up to 9 and 3 km), and one area of rolling hills in Northeast Brazil (up to 9 km). For this, an extensive ensemble of 32 simulations over two continuous years was conducted. Including the reference uncertainty of three high-resolution global datasets (CHIRPS, MSWEP, ERA5-Land), we show that different parameterization setups can produce up to four times the monthly reference precipitation. This underscores the urgent need to conduct parameterization sensitivity studies before weather forecasts or input for impact modeling can be produced. Contrarily to usual studies, we focus on distributional, temporal and spatial precipitation patterns and evaluate these in an ensemble-tailored approach. These ensemble characteristics such as ensemble Structure-, Amplitude-, and Location-error, allow us to generalize the impacts of combining one parameterization scheme with others. We find that varying the CU and RA schemes stronger affects the WRF performance than varying the MP or PBL schemes. This effect is even present in the convection-resolving 3-km-domain over Ecuador/Peru where CU schemes are only used in the parent domain of the one-way nesting approach. The G3D CU physics ensemble best represents the CHIRPS probability distribution in the 9-km-domains. However, spatial and temporal patterns of CHIRPS are best captured by Tiedtke or BMJ CU schemes. Ecuadorian station data in the 3-km-domain is best simulated by the ensemble whose parent domains use the KF CU scheme. Accounting for all evaluation metrics, no general-purpose setup could be identified, but suited parameterizations can be narrowed down according to final application needs.

Evaluation of the WRF Model to Simulate a High-Intensity Rainfall Event over Kampala, Uganda

Simulating high-intensity rainfall events that trigger local floods using a Numerical Weather Prediction model is challenging as rain-bearing systems are highly complex and localized. In this study, we analyze the performance of the Weather Research and Forecasting (WRF) model’s capability in simulating a high-intensity rainfall event using a variety of parameterization combinations over the Kampala catchment, Uganda. The study uses the high-intensity rainfall event that caused the local flood hazard on 25 June 2012 as a case study. The model capability to simulate the high-intensity rainfall event is performed for 24 simulations with a different combination of eight microphysics (MP), four cumulus (CP), and three planetary boundary layer (PBL) schemes. The model results are evaluated in terms of the total 24-h rainfall amount and its temporal and spatial distributions over the Kampala catchment using the Technique for Order of Preference by Similarity to Ideal Solution (TOPSIS) analysis. Rainfall observations from two gauging stations and the CHIRPS satellite product served as benchmark. Based on the TOPSIS analysis, we find that the most successful combination consists of complex microphysics such as the Morrison 2-moment scheme combined with Grell-Freitas (GF) and ACM2 PBL with a good TOPSIS score. However, the WRF performance to simulate a high-intensity rainfall event that has triggered the local flood in parts of the catchment seems weak (i.e., 0.5, where the ideal score is 1). Although there is high spatial variability of the event with the high-intensity rainfall event triggering the localized floods simulated only in a few pockets of the catchment, it is remarkable to see that WRF is capable of producing this kind of event in the neighborhood of Kampala. This study confirms that the capability of the WRF model in producing high-intensity tropical rain events depends on the proper choice of parametrization combinations.

Investigating the efficacy of a new symmetric index of agreement for evaluating WRF simulated summer monsoon rainfall over northeast India

The efficacy of the standard performance metrics [mean bias (Bias), root mean square deviation (RMSD), and correlation coefficient (CC)] compared to a new symmetric index of agreement (λ) for the evaluation of numerical weather prediction models is investigated in this study. It evaluates the weather research and forecasting (WRF) model with the global precipitation measurement’s (GPM) Integrated Multi-satellitE Retrievals for GPM (IMERG) and station rainfall, as the reference datasets. This study uses three IMERG products, namely: GPM infrared-microwave merged gauge corrected (GMS), GPM microwave-calibrated infrared (GIR), and GPM inter-calibrated microwave (GMW) rainfall. The analysis showed that WRF rainfall when compared to different reference datasets is producing similar RMSD values but significantly different Bias values. This behavior is because of the inverse relationship between Bias and standard deviation of residual ( $$\sigma_{{\text{R}}}$$ ). It is so because RMSD is a function of both. However, λ is able to appropriately represent the distinct performances of WRF. The regions with contradictory behavior of RMSD and CC are also appropriately represented in λ. The evaluation using λ showed that WRF is comparable to GMS and GIR, except for GMW. The performance of WRF was not found to be very promising when compared to station rainfall, which is attributed to WRFs representation efficiency and the effect of topography. However, a comparison of IMERG products with station rainfall showed that GMS was the most agreeable followed by GIR and GMW. The study also showed that the efficacy of λ is related to its non-linear relationship with Bias and CC.

Satellite Observation for Evaluating Cloud Properties of the Microphysical Schemes in Weather Research and Forecasting Simulation: A Case Study of the Mei-Yu Front Precipitation System

Radiative transfer model can be used to convert the geophysical variables (e.g., atmospheric thermodynamic state) to the radiation field. In this study, the Community Radiative Transfer Model (CRTM) is used to connect regional Weather Research and Forecasting (WRF) model outputs and satellite observations. A heavy rainfall event caused by the Mei-Yu front on the June 1, 2017, in the vicinity of Taiwan, was chosen as a case study. The simulated cloud performance of WRF with four microphysics schemes (i.e., Goddard (GCE), WRF single-moment 6 class (WSM), WRF double-moment 6 class (WDM), and Morrison (MOR) schemes) was investigated objectively using multichannel observed satellite radiances from a Japanese geostationary satellite Himawari-8. The results over the East Asia domain (9 km) illustrate that all four microphysics schemes overestimate cloudy pixels, in particular, the high cloud of simulation with MOR when comparing with satellite data. Sensitivity tests reveal that the excess condensation of ice at ≥14 km with MOR might be associated with the overestimated high cloud cover. However, GCE displayed an improved performance on water vapor channel in clear skies. When focusing on Taiwan using a higher (3 km) model resolution, each scheme displayed a decent performance on cloudy pixels. In the grid-by-grid skill score analysis, the distribution of high clouds was the most accurate among the three cloud types. The results also suggested that all schemes required a longer simulation time to describe the low cloud horizontal extend.

On the Analysis of the Performance of WRF and NICAM in a Hyperarid Environment

Abstract The Weather Research and Forecasting (WRF) Model and the Nonhydrostatic Icosahedral Atmospheric Model (NICAM) are forced with the Global Forecast System (GFS) data and run over the United Arab Emirates (UAE) for two 4-day periods: one in the cold season (16–18 December 2017) and another in the warm season (13–15 April 2018). The models’ performance is evaluated against four observational datasets: weather station observations, eddy-covariance flux measurements at Al Ain, microwave radiometer–derived temperature profile, and twice-daily radiosonde measurements at Abu Dhabi. An overestimation of the daily mean air temperature by 1°–3°C is noticed for both models and periods. This warm bias is attributed to the reduced cloud cover and resulting increased surface downward shortwave radiation flux. A comparison with the eddy-covariance data suggested that both models also underestimate the observed albedo. However, when the models predict heavier amounts of precipitation, they tend to be colder than observations, typically by 2°–3°C. NICAM and WRF overpredict the strength of the near-surface wind speed at all weather stations by roughly 1–3 m s−1, which has been attributed to a poor representation of its subgrid-scale fluctuations and surface drag parameterization. WRF tends to be wetter and NICAM drier than the station observations, possibly because of differences in the cloud microphysics schemes. While the performance of both models for the near-surface fields is comparable, NICAM outperforms WRF in the simulation of vertical profiles of temperature, relative humidity, and wind speed, being able to partially correct some of the biases in the GFS data.

Influence of land use on the performance of the WRF model in a humid tropical climate

This study evaluates the effect of the land use update to meteorological simulations in the Metropolitan Region of Greater Vitoria (MRGV) using the Weather Research and Forecasting (WRF) mesoscale model. WRF performance was analyzed using monitoring data like wind speed and direction, temperature, and sensible heat flux. In general, the changes in the simulations results with the updating of the land use data are quite subtle, and these values are better represented by the local data in the cold front period. The results obtained indicate a slight improvement in the wind speed and direction when land use is local for the airport station, where the mean bias reduced from − 0.124 to − 0.079, thereby better capturing smaller urban centers and the urban canopy.

Simulation of Sea Breeze Events in Gulf of Jakarta under Different Synoptic Condition: An Application of WRF Model

The behaviour of the sea breeze developing along the Gulf of Jakarta during days with different synoptic condition is examined using observational data and simulations from the Weather Research and Forecasting (WRF). An objective and systematic selection method is used to detect sea breeze days from observational data. The simulation used three nested grids at 27, 9 and 3 km resolution covering Gulf of Jakarta and its joining region for sea breeze event at November 22th, 2016 which formed under offshore large-scale ambient wind, January 20th, 2017 which formed under westerly large-scale ambient wind, and August 10th, 2017 which formed under easterly large-scale ambient wind. Results show that the development of sea breeze circulation under offshore ambient wind produced lower 10-m wind speed simulation, while sea breeze conditions under westerly large-scale ambient wind are associated with highest surface wind speeds reaching 9.5 ms-1 in Gulf of Jakarta. Sea breeze simulation also indicates the formation of calm zone which is 22 - 56 km away from the Jakarta Gulf coastline. Model also shows the formation of thermal internal boundary layer that are about 300-700 m near the coast gradually merged with the mixing layer inland. Performance of WRF on the simulated circulation is studied by conducting value of correlation, RMSE, and MBE. Overall, model simulation agrees fairly well with the observations taken at a near-surface meteorological station.

Assessing Bias Correction Methods in Support of Operational Weather Forecast in Arid Environment

In this study, the Weather Research and Forecasting (WRF) model is employed for operational forecasting over the United Arab Emirates (UAE). The goal of this study is to assess two bias correction methods, namely the multiplicative Ratio Correction (RC) and Kalman Filter (KF), in support of operational mesoscale forecasts in the UAE. These techniques are applied to the 2-m temperature with the corrected temperature subsequently used to update the Relative Humidity (RH) predictions. The simulation covers the 2-year period 1st January 2017 to 31st December 2018. To evaluate the WRF performance, Meteorological Aerodrome Reports (METARs) observations at five airport stations are used. It is concluded that when any of the bias correction techniques are applied, there is a significant reduction of the bias and Root-Mean-Square-Error (RMSE). This is particularly true in the summer season and during nighttime and early morning hours, when WRF has a systematic cold bias of up to 2 °C. In addition, the bias distribution is more symmetric with a reduced spread, skewness and kurtosis values. The RC technique is found to give the best scores, with the observed and modelled temperatures generally within 0.25 °C for the first two forecast days. In addition, it successfully removes the model tendency of underperforming in the warm season. A similar improvement in the skill scores is seen in the RH forecasts albeit with smaller magnitudes. The KF and RC techniques used here have been employed successfully in operational forecasts with the potential to expand them to other model variables.

Evaluation of WRF Modeling in Relation to Different Land Surface Schemes and Initial and Boundary Conditions: A Snow Event Simulation Over the Tibetan Plateau

AbstractSnowfall and the subsequent evolution of the snowpack play important roles in the cryospheric and hydrospheric processes that occur on the Tibetan Plateau (TP). Current literature provides scarce evidence covering the sensitivity of solid precipitation to land surface physics schemes and initial and boundary conditions on the TP. Six numerical experiments using the Weather Research and Forecasting (WRF) model were conducted to simulate a snow event over the TP in March 2017. Different land surface physics schemes, that is, Community Land Model (CLM), Noah, and Noah‐MP, and initial and boundary conditions provided by atmospheric reanalysis data sets, that is, the National Centers for Environmental Prediction‐FNL and ERA‐Interim data sets, were applied in sensitivity analyses. The observed near‐surface air temperature, snow depth, and snow water equivalent (SWE) values were used to evaluate each model's performance. The results demonstrate that (1) the sensitivity of the near‐surface air temperature to land surface physics schemes is greater than it is to both the initial and boundary conditions; (2) the best performance is achieved when applying WRF + CLM with a root‐mean‐square error of 8.4 °C, a mean absolute deviation of 7.3 °C, a correlation coefficient of 0.75, and a spatial correlation coefficient of ~0.5 to air temperature estimates. A potentially important factor appears to be the advanced parametrization of albedo in the CLM scheme; (3) the advanced land surface schemes in the WRF model describes the physics of cryospheric and hydrospheric processes in detail, and the land surface response is determined by multiple variables and parameters in such schemes. The spatial patterns in such variables and parameters determined the detailed spatial variabilities observed in snow cover and amount and its temporal evolution. The WRF model overestimates, however, the intensity and extent of snow depth and SWE; (4) simulations of solid precipitation are more accurate when applying CLM or Noah‐MP + ERA‐Interim in WRF; and (5) WRF performance with regard to SWE estimates clearly depends upon the discrimination of lighter from heavier snowfall.

The Impact of Spatial Resolution, Land Use, and Spinup Time on Resolving Spatial Precipitation Patterns in the Himalayas

Abstract Frequently used gridded meteorological datasets poorly represent precipitation in the Himalayas because of their relatively low spatial resolution and the associated representation of the complex topography. Dynamical downscaling using high-resolution atmospheric models may improve the accuracy and quality of the precipitation fields. However, most physical parameterization schemes are designed for a spatial resolution coarser than 1 km. In this study the Weather Research and Forecasting (WRF) Model is used to determine which resolution is required to most accurately simulate monsoon and winter precipitation, 2-m temperature, and wind fields in the Nepalese Himalayas. Four model nests are set up with spatial resolutions of 25, 5, 1, and 0.5 km, respectively, and a typical 10-day period in summer and winter in 2014 are simulated. The model output is compared with observational data obtained from automatic weather stations, pluviometers, and tipping buckets in the Langtang catchment. Results show that, despite issues with the quality of the observational data due to undercatch of snowfall, the highest resolution of 500 m does provide the best match with the observations and gives the most plausible spatial distribution of precipitation. The quality of the wind and temperature fields is also improved, whereby the cold temperature bias is decreased. Our results further elucidate the performance of WRF at high resolution and demonstrate the importance of accurate surface boundary conditions and spinup time for simulating precipitation. Furthermore, they suggest that future modeling studies of High Mountain Asia should consider a subkilometer grid for accurately estimating local meteorological variability.

Assessing the performance of WRF model in predicting high-impact weather conditions over Central and Western Africa: an ensemble-based approach

For numerical weather prediction over a particular region, it is important to know the best combination of physical parameterizations available in the considered modelling frame work. The main objective of the study is to obtain the best combination of the weather research and forecasting (WRF) model physics for accurately simulating high-impact weather conditions, especially the rainfall over Western and Central Africa. For this purpose, performance of WRF from various simulations with specific configurations is assessed by comparing the results to the Tropical Rainfall Measurement Mission data. Each of the simulations is carried out for 30 h and initialized at 00 UTC. The spin-up time considered for the study is 6 h. A flood event (21–22 July 2010) is simulated by considering five cumulus physics schemes with multiple cloud microphysics, planetary boundary layer and land surface parameterizations. Analysis of the model results indicates that some of the physics combinations have good agreement with observations, especially the new (GFS) simplified Arakawa–Schubert and the modified Tiedtke cumulus parameterizations combined with Thompson and Morrison microphysics. However, most of the combinations over-estimated the rainfall over the study domain, while the simulations with Betts–Miller–Janjic cumulus parameterizations showed negative bias over designated regions of Africa.