Next Article in Journal
Particle Number Size Distribution in Three Different Microenvironments of London
Previous Article in Journal
Air Pollution and Climate Change Risk Perception among Residents in Three Cities of the Mexico Megalopolis
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Atmosphere
Volume 15
Issue 1
10.3390/atmos15010044
atmosphere-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles thumb_up
...
Endorse textsms
...
Comment
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Precipitation Extremes over India in a Coupled Land–Atmosphere Regional Climate Model: Influence of the Land Surface Model and Domain Extent

by
Alok Kumar Mishra
1,2,*,
Anand Singh Dinesh
3,
Amita Kumari
4 and
Lokesh Kumar Pandey
3
1
K Banerjee Centre of Atmospheric and Ocean Studies, University of Allahabad, Prayagraj 211002, India
2
Volcani Institute, Agricultural Research Organization, Rishon LeZion 7505101, Israel
3
Department of Earth and Environmental Sciences, Indian Institute of Science Education and Research Bhopal, Bhopal 462066, India
4
The Fredy & Nadine Herrmann Institute of Earth Sciences, Hebrew University Jerusalem, The Edmond J. Safra Campus—Givat Ram, Jerusalem 9190401, Israel
*
Author to whom correspondence should be addressed.
Atmosphere 2024, 15(1), 44; https://doi.org/10.3390/atmos15010044
Submission received: 27 September 2023 / Revised: 3 December 2023 / Accepted: 5 December 2023 / Published: 29 December 2023
(This article belongs to the Section Atmospheric Techniques, Instruments, and Modeling)
Browse Figures
Versions Notes

Abstract

:
The frequency and intensity of extreme precipitation events are on the rise worldwide. Despite extensive efforts, regional climate models still show significant biases for extreme precipitation events, often due to factors like improper physics, the choice of land surface model, and spatial domain. Thus, this study uses a Coupled Land–Atmosphere Regional Climate Model version 4.7 (RegCM4.7) to explore how the choice of land surface models (LSMs) and domain extent affects the simulation of extreme precipitation over India. In this regard, a total of four sensitivity experiments have been carried out using two LSMs (CLM4.5 and BATS) over each of the two domains (one over the bigger South Asia CORDEX domain and another for the smaller domain over the Indian region). The main objective is to provide a holistic idea for obtaining an optimum model domain and LSMs for precipitation extremes over India. The model performance is demonstrated for extreme precipitation and associated processes. The result shows the systematic discrepancy in simulating extreme precipitation with a strong inter-simulation spread, indicating the strong sensitivity of extreme precipitation on the LSMs as well as the model domain. The BATS configuration shows a significant overestimation of consecutive wet days and very low precipitation, partially associated with a deficiency in convection. By contrast, the considerable underestimation of intense precipitation can be attributed to the presence of frequent, light drizzle, which hinders the accumulation of moisture in the atmosphere to a sufficient degree to prevent extreme rainfall. Despite significant improvement, the best-configured model (CLM with Indian domain) still indicates substantial bias for extreme precipitation. This deficiency in the model could potentially be mitigated by enhancing both horizontal and vertical resolutions. Nevertheless, further research is needed to explore other physics parameterizations and dynamic mechanisms to address this issue.

1. Introduction

India is one of the most vulnerable regions among the South Asian countries and is highly affected by high-frequency extreme climate events. This includes intense high and low rainfall (flood and drought) and prolonged heat waves and cold waves. All these events adversely affect the hydrological cycle, agriculture, and the lives of people [1,2,3]. These extreme events have profound impacts on the understanding of water hydrological dynamics for the efficient distribution, accessibility, and sustainable stewardship of freshwater resources for the proper functioning of a multitude of ecosystems [4,5]. The crop yield accuracy (deficiency) simulated by crop models is dependent on the precision of input fields such as temperature or rainfall. Consequently, crop production is directly influenced by both the regular conditions of precipitation and temperature, as well as instances of their extreme events [6]. Therefore, the timely and accurate prediction of extreme rainfall and temperature events is important. Although temperature modeling is well-addressed, issues in producing rainfall patterns with several levels of complexity persist. As a result, better rainfall estimation remains an enticing goal for the climate model community.
It is being reported that climate variability over the South Asia region has dramatically changed in recent decades, with considerable heterogeneity in space and time. Previous studies have reported a weakening of the mean Indian summer monsoon rainfall and a strengthening in the intensity and frequency of the extreme precipitation events over India in the warming scenarios using different climate models [7,8,9,10]. However, the projections from the climate models (regional or global) are generally associated with a broad range of limitations and uncertainties [11]. Therefore, the precise prediction of such extreme events remains challenging for the scientific community. This is most likely due to the coarse resolution of Global Climate Models (GCMs), which have limitations in resolving the mesoscale processes that significantly contributed to extreme events, as well as the poor understanding of the mesoscale processes responsible for the extreme events, which is also due to the uncertainties. Thus, identifying the source of uncertainty and correctly comprehending the factors responsible for the extremes is essential [12]. According to several CMIP5 model-based studies, the uncertainty in the temperature over most regions of India is caused by model spread rather than natural internal variability and trends. The authors in [13] demonstrated the performance of the different CMIP5-GCMs in simulating the maximum and minimum temperatures over India. A substantial cold (warm) bias in minimum (maximum) temperature has been reported. This necessitates a thorough explanation of the model and further explains how the tuning or tweaking of the different components of the model physics can be helpful in selecting the best model setup before using it for future projection.
The high-resolution regional climate models (RCMs) have the advantage of resolving the coastlines and topography [14,15,16,17,18] and land surface feedback processes, which helps to improve the simulation of both large-scale as well as small-scale extreme weather events [19,20,21,22,23,24,25]. Consequently, RCMs capture extreme events in a much better fashion than coarse-resolution GCMs [7,21,26,27]. The majority of the RCM studies report the advantage of increasing the model’s resolution [16,17,18]; however, merely increasing the resolution blindly without additional tuning is not sufficient [16,17]. Studies have further highlighted the benefits of tuning and optimizing model setups in terms of convective parameterization schemes, land surface models (LSMs; here onwards referred to interchangeably as land surface schemes or LSSs), boundary conditions, microphysical schemes, and domain [27,28,29,30,31,32,33].
Some studies have reported improved performance by the incorporation of complex LSMs, namely CLM4.5, which has led to improvements in simulating the mean monsoon rainfall over India [33,34]. Conversely, other research has suggested that the adoption of advanced land surface schemes may not consistently yield enhanced performance [29]. Thus, additional research is required to improve our understanding of land–atmosphere feedback and determine the suitable LSMs, especially in the context of regional precipitation extremes, which have been little explored.
Moreover, the quality of simulated fields is significantly influenced by the domain’s extent, as highlighted by previous research [35,36,37]. Some studies have recommended conducting simulations with an optimized domain size, emphasizing the importance of a larger domain to capture small-scale features within the target area [36]. The authors in [28] noted improved performance in simulating mean monsoon rainfall when using a larger domain compared to a smaller one. Nevertheless, an excessively large domain can lead to discrepancies in reproducing the large-scale features derived from the driving model [37]. In contrast, recent studies by [16,17,30] have shown improved results with a smaller domain than the larger South Asia CORDEX domain (here onwards interchangeably referred to as SACD or SA) for simulating Indian monsoon rainfall.
The above review distinctly states that the simulation of monsoon precipitation is sensitive to both the choice of domain and LSMs. Therefore, in this study, we revisit this aspect to explore the possible impact of the LSMs available in RegCM4.7, as well as the domain extent of the simulation of extreme precipitation over India. To the best of the authors’ knowledge, no specific studies have delved into the combined influence of LSMs and domain extent on regional precipitation extremes over India. Additionally, we explore how the performance of a particular LSM depends on domain extent. The subsequent sections provide detailed descriptions of the model, results, and conclusions.

2. Model and Methodology

2.1. Model Setup and Experimental Design

This study utilizes the Regional Climate version 4.7 (RegCM4.7) developed by the authors in [38] to investigate how the choice of LSMs and domain extent affects the simulation of precipitation extremes. To achieve this, four experiments were conducted. Two of these experiments focused on different LSMs, specifically the Biosphere–Atmosphere Transfer Scheme (BATS) [39] and the Community Land Model version 4.5 (CLM4.5) [40]. The other two experiments explored the impact of domain extent.
The initial and lateral boundary conditions for all simulations are derived from the European Center for Medium-Range Weather Forecasting (ECMWF) ERA-Interim reanalysis data, which is available at 6-h intervals [41]. Weekly sea-surface temperature, with a horizontal resolution of 1.0° × 1.0°, is derived from the National Oceanic and Atmospheric Administration’s Optimum Interpolation Sea-Surface Temperature [42]. The land use data and terrain heights are generated from the United States Geographical Survey at a 30-s resolution. All simulations were carried out for 11 years, from 1 January 2005 to 31 December 2015. The horizontal resolution is 50 km, and 23 vertical levels are used within the Sigma coordinate. The analysis presented in the paper focuses on the summer monsoon season (June–July–August–September; JJAS) across ten years. The first 17 months are excluded as a spin-up time to remove initial transients. Details of the experiments are provided in Table 1 and Figure 1.
Although a ten-year period is not a long simulation to produce robust statistics, it does allow for the preliminary demonstration of the possible influence of the LSMs and domain extent. This can provide direction for selecting possibly the most computationally efficient setup, which can be utilized for longer simulations and further tuning to improve the performance.
A daily gridded high-resolution 0.25° × 0.25° precipitation dataset from the India Meteorological Department (IMD) [43] is used as an observation to demonstrate model potential in simulating precipitation extremes. Additionally, convective and large-scale precipitation, specific humidity, and circulation at 850 hPa data have been used from the fifth generation of ECMWF Global Reanalysis (ERA-5) products [44].

2.2. Study Areas (Model Domains)

The larger domain simulation is performed over the South Asia Coordinated Regional Climate Downscaling Experiment (CORDEX) domain (hereafter, this domain is referred to as SACD) defined by the authors in [45]. CORDEX is an international effort aimed at producing high-resolution climate projections for various regions around the world. South Asia is one of the regions where CORDEX has established a firm domain to conduct regional climate modeling and downscaling experiments. This domain encompasses a substantial number of South Asian countries, including India, Pakistan, Bangladesh, Nepal, Bhutan, and Sri Lanka. It covers a diverse landscape, including high mountain ranges, vast plains, and coastal regions, such as the Himalayan Mountain Range, the Indo–Gangetic Plain, the Western and Eastern Ghats, and coastline along the Indian Ocean, with various coastal features such as beaches, deltas, and estuaries, arid regions, and plateaus. This diversity makes the simulation of the region very challenging. This domain allows climate scientists to focus on the specific climatic conditions and changes in South Asia.
The smaller domain focuses on a subset of the SACD area, covering only the Indian region (hereafter, this domain is referred to as IND), spanning from 45° E to 110° E longitude and from 10° S to 41° N latitude, as shown in Figure 1.

2.3. Extreme Precipitation Indices

This study employed a set of extreme indices developed by the Expert Team on Climate Change Detection and Indices (ETCCDI) to investigate its dependency on LSMs and domain extent. The spatial distribution of the percentile-based precipitation indices for heavy extremes (that represent the amount of rainfall, in mm/day, falling above 95th (R95) and 99th (R99) percentiles) and low or shallow precipitation events (50th (R50) and 10th (R10) percentiles) is computed. Additionally, consecutive dry days (CDD) (the number of consecutive days per time period with a daily precipitation amount below 1 mm) and consecutive wet days (CWD) (the number of consecutive days with a daily precipitation amount above 1 mm) are also computed. These indices are calculated for all simulations and observations (IMD) over India for JJAS months covering 2006–2015 and have been used previously in many studies [7,46,47]. Apart from this, the probability density function (PDF) [48] is estimated to showcase the overall distribution of frequency and intensity over six Indian homogeneous rainfall regions (IHRR), where the PDF tails represent the precipitation extremes with their probability of occurrence over time.

3. Results

3.1. Comparative Evaluation for Precipitation Extremes

This study aims to illustrate the sensitivity of extreme precipitation to the LSMs and domain extent in RegCM4.7 over India. However, it is worth demonstrating how the model simulation represents the mean precipitation and its variability over different homogeneous regions. Thus, we computed the daily climatology of precipitation during JJAS months for all the simulations and observations (IMD); see Figure S1. All simulations show reasonable skills in reproducing the mean precipitation imprints; however, magnitudes show noticeable spreads among the simulations that also vary with regions. For example, BATS/CLM (with both domains) shows underestimation/overestimation over northeast India (NEI), particularly during the onset phases; however, CLM is closer to observation. Apart from this, it is also noticeable that the difference due to the land surface schemes (LSSs) is larger compared to the difference due to the domain throughout the season. However, over CNE, both LSSs and domain extent show comparable spread.
To make the comparison more robust, we also computed the correlation coefficient between model-simulated precipitation and IMD precipitation (Table S1), which shows that the smaller domain (IND) has a higher correlation than the larger domain (SACD) over all the homogeneous regions except NEI. The influence of the LSSs is found to have a strong dependency on the domain. For example, over CNE, the differences in the correlation values between BATS and CLM are smaller for the IND domain than for the SACD domain. A similar inference is also noticed for the mean precipitation (Table S1).
We made a comparative evaluation of all simulations for the spatial distribution of precipitation (low to heavy) in terms of the percentage of precipitation contribution. Figure 2 represents the extreme precipitation (R95 and R99) in all simulations and observations. The figure shows the substantial heterogeneity in extreme precipitation (R95) and very extreme precipitation regimes (R99). In the observation (IMD), extreme precipitation (R95 and R99) is observed over central India, some parts of NEI, western ghats, and northwest India (NWI). Given that the NWI region is known for receiving much less precipitation, both in terms of intensity and frequency, than the rest of the country, one could wonder about the high values of R99 and R95 over the NWI. It does, however, frequently exceed the R95/R99 in this region. This pattern is consistent with the recently reported study by [49]. All model simulations bear a reasonable resemblance to observation in reproducing this prominent pattern of the spatial variability of R95 and R99 over India. However, the magnitude of precipitation intensity shows significant differences relative to observation and varies substantially in all simulations. In general, it is noticed that all simulation shows higher deviation from observation (limited skill) in very high extremes (R99) than in high extremes (R95) and minimal deviation in simulating low (R10) and moderate (R50) precipitation events (Figure 3). It is interesting to note that, unlike the underestimation of high and very high precipitation extremes (R95 and R99) in all simulations relative to observation, the low precipitation intensity (R50 and R10) is found to be overestimated. This discrepancy was substantially reduced by complex land surface schemes and further by reducing the domain. However, the gaps cannot be closed, necessitating additional advancements. For example, several studies reported that extreme precipitation is largely regulated by mesoscale scale activities (small-scale processes) [50,51,52,53,54], demanding further enhancement in the model’s resolution.
Although all simulations show limited skill, the CLM simulation performs relatively better than BATS for both domains in reproducing the extreme precipitation signal over the complex topography and local land surface feedback mechanisms. On the other hand, model simulations over the IND domain outperformed the larger SACD domain for both LSMs. Overall, CLM with the IND domain shows the highest skill in simulating the R95 and R99.
To make results more prospective, the percentage bias for all types of precipitation was computed (Figure 4 and Figure 5) to estimate the quantitative differences. The figure reveals a large spread among simulations in terms of the percentage bias. In general, R95 and R99 show underestimation (10–90%) at most parts of the Indian land region except for some patches over the hilly region (HR), NEI, and southeast central India, where overestimation is noticed. In contrast, R10 and R50 show underestimation (up to 20%) over India. Overall, the model exhibits greater bias (~20–40%) in very high extremes compared to the low-precipitation events, indicating that it is more effective at simulating low-intensity precipitation than high-intensity precipitation. It is evident from the figure that CLM simulation shows lesser dry bias (~10–40%) in simulating R95 and R99 over the core monsoon zone and more significant wet bias (up to 10%) over some parts of the hilly region compared to the BATS simulation, irrespective of the domains. Furthermore, lesser dry bias (up to 20%) is noticed in the IND domain simulation over the core monsoon zone, and larger wet (dry) bias (5 to 20%) for R95 and R99 over some parts of the hilly region (southern peninsular India; SPI) compared to bigger SACD domain simulation in LSMs. Apart from this, some places have a bias of a similar nature and approximately comparable magnitude, indicating that both domain and LSMs may not be playing a significant role in the bias over these regions.
For R10 and R50, BATSSA simulation shows the wet bias of (5–15%) over SPI and the Indo–Gangetic plain. The smaller domain (BATSIND) substantially reduces the bias over these regions but slightly increases over central India. The incorporation of CLM further reduces bias compared to BATS for both domains. This detailed investigation further boosts confidence in the increased proficiency of the CLM simulation in combination with the IND domain (CLMIND).
The discussion indicates an overestimation of lower-intensity precipitation and an underestimation of higher-intensity precipitation, which generally compensates for the hidden error in the mean state evaluation. Thus, it is worthwhile to determine the model’s skill for a range of frequencies and intensities that will help to isolate and determine the origin of the hidden uncertainty in the mean state. In this regard, the PDFs of the daily mean precipitation for the summer monsoon season over six IHRRs were computed (Figure 6). It is noted from the figure that all simulations bear a close resemblance to the observation of the low-intensity precipitation events over all regions. The BATS simulations significantly underestimate the occurrence of moderate to very high-intensity precipitation events (>100 mm/day), particularly over NWI, CNI, and WCI, with a varying magnitude of underestimation over the different regions as indicated by the longer tails. In general, the highest spread among simulations is noticed over WCI in terms of underestimation, and the observed intensity tail is relatively well reproduced over SPI, NEI, and HR. The incorporation of CLM shows the tendency to produce high-intensity precipitation, reducing the underestimations over most regions. However, it slightly overestimates the regions with a complex topography (HR and NEI). This might be due to considerable uncertainty in measurements caused by a very sparse network of IMD rain gauge stations over these regions. Overall, CLMIND shows better performance among all in simulating precipitation intensity, particularly for high intensity. Although there has been considerable enhancement with CLM, the simulation of extreme precipitation intensity still has not attained a satisfactory level. It is noteworthy that the bias in this study is substantially less than earlier studies using the multimodel ensemble mean of CMIP, which reported the failure of the ensemble mean in simulating the intensity of heavy extreme events, with biases even 4–5 times smaller than observation [55] which remains challenging for the modeling community. Climate scientists continue to implement various strategies, including improving model physics to improve performance.
Furthermore, in all simulations, spatial patterns of CDD and CWD were compared with the corresponding observed values. In this regard, bias is computed for JJAS mean CDD and CWD (Figure 7). The figure reveals strong sensitivity of CDD and CWD to both LSMs and domain extent. However, the impact of LSMs is found to be slightly larger than the domains. It can be noted from the figure that the BATSSA simulation shows substantial positive bias in simulating CWD over most parts of India, with maximum bias over east-central India.
On the other hand, a slight negative bias is noticed over NWI and the Himalayan region. A similar nature of bias is also noted in the BATSIND simulation over most parts of India except the eastern part of SPI, where it shows a negative bias in contrast to a positive bias in BATSSA. However, the magnitude of bias is substantially reduced in BATSIND compared to BATSSA. It is interesting to note that, unlike switching from a larger domain (SACD) to a smaller domain (IND), changing LSMs from BATS to CLM produces a contrasting nature of bias. For example, BATS shows a strong positive bias over CI, whereas CLM shows minimal negative bias over the same region. In BATS over CI, despite the overestimation of CWD, extreme precipitation (R95 and R99) shows high underestimation, indicating the possibility of continuous, very low-intensity precipitation events, which is also partially supported by the overestimation of R10 and R50. This could be partly associated with the deficiency in triggering the convection mechanism. Once stabilization is achieved, it may persist even in the presence of actual destabilization. Additionally, it could be attributed to the presence of high-frequency, low-intensity drizzle that prevents the moisture from building up in the atmosphere to a level leading to the underestimation of extreme events. The comparison of CDD reveals an underestimation in all simulations. Nevertheless, CLM exhibits a lesser bias than BATS, consistent with corresponding improvements in precipitation amount.

3.2. Physical Mechanism Associated with Extreme Precipitation in Different Simulations

Numerous processes (i.e., mesoscale convective process, cloud bursts, cyclones, excess moisture availability, favorable surface conditions, and geographical attributes such as topography) contribute to extreme precipitation events. In the model, the convective processes mainly produce convective precipitation (CP), while sub-grid-scale precipitation or large-scale precipitation (LSP) are governed by large-scale dynamics. The model’s potential to simulate extreme precipitation hinges on accurately representing convection processes and appropriately partitioning CP and LSP. However, their proper simulation remains a significant challenge. Therefore, diagnosing the model’s sensitivity to reproduce the CP and LSP during extreme precipitation events is worthwhile. Figure 8 represents the daily composite maps of CP obtained from the ERA5 [44] during extreme precipitation days. Days with precipitation exceeding R95 in central India are considered extreme rainfall events. The figure suggests a predominant contribution of convective precipitation over central India, while LSP is primarily observed over the Western Ghats and Western Central India.
All simulations show that CP dominates LSP over central India during extreme rainfall events (Figure 9 and Figure 10). However, the magnitude and area vary among the different simulations. Figure 9 reveals that all simulations show a similar spatially structured CP anomaly, i.e., an increase (decrease) in convective precipitation over central India (the Gangetic Plain). However, a noticeable difference is noted in terms of the size of the anomaly across each simulation. Generally, the CLM exhibits a more substantial increase (decrease), a positive (negative) of more than 10 mm/day over central India (the Gangetic plain) compared to BATS for both domains. The magnitude of the anomalies is similar in both the domain simulations for the corresponding LSM, with a slight difference in terms of the area of the positive (negative) anomalies. The SACD simulations show a larger extent of positive anomaly over central India and the northern Bay of Bengal (BoB). This indicates that CP is sensitive to the LSMs as well as domain extent. The CP is found to contribute strongly to extreme precipitation in all simulations. Barring a few patches, the areas with higher (lower) convective precipitation in respective simulations are consistent with the regions of high (low) R95 (Figure 2). The CLMIND shows the closest resemblance to ERA5 among all simulations, yet notable differences persist in both magnitude and location. This may be attributed to a combination of factors, including deficiencies in RegCM’s convective parameterization schemes and the large-scale disparities in convection and precipitation production between ERA5 and the RegCM simulation. The production of CP in models is associated with the convective parameterization schemes, and the model’s deficiency in simulating the deep convective processes caused by parameterization-induced inaccuracy for the model grid spacing can lead to uncertainty in the total precipitation. Improving the model’s convective schemes or increasing horizontal resolution can further reduce the uncertainty.
Most of the Indian land region exhibits lesser LSP (Figure 9) than CP (Figure 10), except for some regions along Amrawati. However, both the extent and the intensity of LSP are found to be sensitive to domain and LSMs. Regardless of the domain, CLM simulations typically show an overestimation of LSP over a larger area than the BATS simulations. The larger LSP in CLM, as compared to BATS, may be partly associated with the higher number of soil layers and a better representation of the vegetation cover. A similar inference is reported by the authors in [56,57].
Similarly, IND domain simulation has higher LSP irrespective of LSMs. In particular, the highest LSP is noted in CLMIND, which is collocated with a high R95. The better production of LSP (CP) in the CLMIND improves the permanence of precipitation extremes over Amrawati (central India). The location of higher LSP in the model simulation is not consistent with ERA5. The reason for the contrasting pattern in ERA5 and the model’s simulation is not apparent, indicating it needs further investigation, which is beyond the scope of this study.
The differences in orographic precipitation production among the various simulations may also be attributed to sensitivity to mountain ranges. Consequently, we evaluated elevation-based precipitation. To do so, we divided the elevation into classes of 200 m using the topography of the simulations and computed the precipitation for each class (Figure S2). The figure shows a large difference in the simulated precipitation across different elevation classes. In general, the BATS/CLM notably underestimates and overestimates the observed precipitation in areas with elevations up to 600 m. The IND domain was found to produce larger precipitation than the SACD domain for elevations >600 m and <1600 m. This indicates that LSMs (domain extent) have larger sensitivity for elevations up to 600 m (>600 m and < 1600 m). Interestingly, it is observed that beyond the elevation of 1600 m, LSSs further show larger sensitivity, and CLM with the IND domain produces the highest precipitation.
A comparative analysis of these processes across different experiments can offer insights into the respective spread in performance. In this regard, we examined daily composite maps depicting moisture (specific humidity) and circulation at 850 hPa during days of extreme rainfall events (Figure 11). The ERA5 shows the cyclonic circulation centered over central India, enhancing the upward movement of humid air. This, in turn, promotes cloud formation and consequently leads to precipitation. The southwestern (northeastern) part of the cyclonic structure drives the moisture from the Arabian Sea (AS) and Bay of Bengal (BoB). The increase of specific humidity over central India further contributes to the local moisture availability for precipitation. All simulations show the cyclonic structure as well as positive specific humidity anomalies over central India. However, the center of the cyclonic structure is slightly shifted southeastward. A big difference in terms of the location of the cyclonic structure and specific humidity is among simulations that contribute to moisture availability.
The SACD simulation shows a higher increase in the specific humidity and seems to draw moisture from the AS and BoB. Similarly, the BATS shows a higher increase in specific humidity over central India than CLM. Despite having higher specific humidity in the BATS and SACD simulation, the lower precipitation might be due to unfavorable conditions (stable atmosphere) for precipitation (Figure 12).

4. Conclusions

This study investigates the influence of land surface models (LSMs) and model domains in the context of precipitation extremes using the Coupled Land–Atmosphere Regional Climate Model version 4.7 (RegCM4.7). Within RegCM4.7, there are two distinct land surface models, namely BATS and CLM4.5. The more recent and sophisticated CLM4.5 model, while accurate, demands greater computational resources than the simpler BATS model. The smaller model domain (Indian region domain; IND domain), on the other hand, is less computationally demanding than the larger domain (South Asia CORDEX domain; SACD domain). Therefore, before considering the extended simulation or the use of CLM4.5 for future projections, it is crucial to assess its performance compared to the BATS model.
The study conducted a total of four sensitivity experiments, two for LSMs (CLM and BATS) and two for model domains (IND domain and SACD domain). The objective is to gain a comprehensive understanding of which model domain and LSMs is most suitable for assessing precipitation extremes over India. The overall evaluation of model performance reveals that all simulations tend to underestimate extreme precipitation events. This could be partly related to the inadequate convective parameterizations that might not sufficiently capture the intricacies of complex convective processes, such as the frequent occurrence of low-intensity drizzle/light rain, preventing the buildup of moisture in the atmosphere to a level conducive to extreme precipitation. Additionally, the resolution of the model may not be able to resolve fine-scale complex processes, further leading to introducing uncertainty. Fine-tuning physics parameterizations and using higher-resolution models can help improve the simulation of extreme precipitation events.
The BATS configuration shows a notable overestimation of consecutive wet days and low-intensity precipitation, which can be partly attributed to its shortcomings in representing convection. Despite the higher moisture availability (specific humidity), the underestimation of extremes in the BATS with SACD seems to be related to unfavorable conditions (atmospheric stability) that trigger the convection. Replacing the BATS model with the more complex CLM4.5 and shifting from the SACD domain to the IND domain have both led to the improved spatial distribution of low-to-moderate intensity precipitation events and, to some extent, extreme precipitation. The increase in the extreme precipitation in CLM simulation is attributed to increased instability. Despite significant improvements, the model with the best configuration (CLM with the IND domain) still exhibits substantial bias in predicting extreme precipitation. This deficiency in reproducing convective processes highlights the need for further refinement of the model resolution to enhance its ability to represent such processes accurately.
In summary, this study underscores the complexities involved in modeling precipitation extremes and offers a foundation for future research to improve our understanding of these events. It highlights the need for advanced modeling techniques, a better representation of convective processes, and careful consideration of model domain choices and computationally efficient regional climate setups for studying precipitation extremes in India. This can be useful to enhance our ability to predict and adapt to extreme precipitation events in the face of a changing climate. Improving our understanding and modeling of precipitation extremes has far-reaching socio-economic implications. It enables better preparedness, more effective resource allocation, and enhanced resilience in the face of a changing climate. This, in turn, helps protect lives, reduce economic losses, and promote sustainable development in the face of climate-related challenges.

Supplementary Materials

The following supporting information can be downloaded at https://www.mdpi.com/article/10.3390/atmos15010044/s1, Figure S1: The solid lines represent JJAS daily climatology of the precipitation over the homogeneous rainfall regions for all the simulations and corresponding observations for ten years during 2006–2015. The shaded curve represents the interannual standard deviation of each day during JJAS. The x-axis represents the days during JJAS (1 corresponds to the first of June); Figure S2: Altitudinal distribution of the mean precipitation for all simulations and IMD for JJAS seasons during 2006–2015. The percentage of the total grid points covered by each class is also shown. X-axis represents the elevation class of 200 m; Table S1: Comparison of simulated and observed JJAS daily climatology of the precipitation over the homogeneous rainfall regions.

Author Contributions

A.K.M. conceptualized the study, performed simulations, analyzed data, and wrote the manuscript. A.S.D. conceptualizes the study and performs the formal analysis. A.K. and L.K.P. perform the formal analysis. All the authors reviewed and edited the manuscript. All authors have read and agreed to the published version of the manuscript.

Funding

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Institutional Review Board Statement

Not applicable.

Informed Consent Statement

Not applicable.

Data Availability Statement

The observational datasets used in this study are derived from public resources, and model data will be made available upon request to the corresponding author. The data are not publicly available due to the ongoing research.

Acknowledgments

We thank the anonymous reviewers for their constructive comments and suggestions, which have helped us to improve the overall quality of the paper. The authors would like to thank IMD and ECMWF agencies for making their datasets available free of cost.

Conflicts of Interest

The authors declare that they have no known competing financial interest or personal relationships that could have appeared to influence the work reported in this paper.

References

  1. Krishnamurthy, C.K.B.; Lall, U.; Kwon, H.H. Changing Frequency and Intensity of Rainfall Extremes over India from 1951 to 2003. J. Clim. 2009, 22, 4737–4746. [Google Scholar] [CrossRef]
  2. Rajeevan, M.; Bhate, J.; Jaswal, A.K. Analysis of variability and trends of extreme rainfall events over India using 104 years of gridded daily rainfall data. Geophys. Res. Lett. 2008, 35, L18707. [Google Scholar] [CrossRef]
  3. Thornton, P.K.; Ericksen, P.J.; Herrero, M.; Challinor, A.J. Climate variability and vulnerability to climate change: A review. Glob. Chang. Biol. 2014, 20, 3313–3328. [Google Scholar] [CrossRef] [PubMed]
  4. Usman, M.; Ndehedehe, C.E.; Farah, H.; Ahmad, B.; Wong, Y.; Adeyeri, O.E. Application of a Conceptual Hydrological Model for Streamflow Prediction Using Multi-Source Precipitation Products in a Semi-Arid River Basin. Water 2022, 14, 1260. [Google Scholar] [CrossRef]
  5. Maneechot, L.; Wong, Y.J.; Try, S.; Shimizu, Y.; Bharambe, K.P.; Hanittinan, P.; Ram-Indra, T.; Usman, M. Evaluating the necessity of post-processing techniques on d4PDF data for extreme climate assessment. Environ. Sci. Pollut. Res. 2023, 30, 102531–102546. [Google Scholar] [CrossRef] [PubMed]
  6. Prasanna, V. Impact of monsoon rainfall on the total foodgrain yield over India. J. Earth Syst. Sci. 2014, 123, 1129–1145. [Google Scholar] [CrossRef]
  7. Mishra, A.K.; Dwivedi, S.; Das, S. Role of Arabian Sea warming on the Indian summer monsoon rainfall in a regional climate model. Int. J. Climatol. 2020, 40, 2226–2238. [Google Scholar] [CrossRef]
  8. Roxy, M.K.; Ghosh, S.; Pathak, A.; Athulya, R.; Mujumdar, M.; Murtugudde, R.; Terray, P.; Rajeevan, M. A threefold rise in widespread extreme rain events over central India. Nat. Commun. 2017, 8, 708. [Google Scholar] [CrossRef]
  9. Roxy, M.K.; Ritika, K.; Terray, P.; Masson, S. The Curious Case of Indian Ocean Warming. J. Clim. 2014, 27, 8501–8509. [Google Scholar] [CrossRef]
  10. Roxy, R.K.; Ritika, K.; Terray, P.; Murtugudde, R.; Ashok, K.; Goswami, B. Drying of Indian subcontinent by rapid Indian Ocean warming and a weakening land-sea thermal gradient. Nat. Commun. 2015, 6, 7423. [Google Scholar] [CrossRef]
  11. Chaturvedi, R.K.; Joshi, J.; Jayaraman, M.; Bala, G.; Ravindranath, N.H. Multi-model climate change projections for India under representative concentration pathways. Curr. Sci. 2012, 103, 791–802. [Google Scholar]
  12. Singh, R.; AchutaRao, K. Role of Internal Variability in Surface Temperature and Precipitation Change Uncertainties over India IIT Delhi. In Proceedings of the AGU FALL Meet, New Orleans, LA, USA, 11–15 December 2017. [Google Scholar] [CrossRef]
  13. Panjwani, S.; Naresh Kumar, S.; Ahuja, L.; Islam, A.; Panjwani, S.; Naresh Kumar, S.; Ahuja, L.; Islam, A. Evaluation of selected global climate models for extreme temperature events over India. Theor. Appl. Climatol. 2020, 140, 731–738. [Google Scholar] [CrossRef]
  14. Chan, S.C.; Kendon, E.J.; Fowler, H.J.; Blenkinsop, S.; Ferro, C.A.T.; Stephenson, D.B. Does increasing the spatial resolution of a regional climate model improve the simulated daily precipitation? Clim. Dyn. 2013, 41, 1475–1495. [Google Scholar] [CrossRef]
  15. Kripalani, R.H.; Oh, J.H.; Kulkarni, A.; Sabade, S.S.; Chaudhari, H.S. South Asian summer monsoon precipitation variability: Coupled climate model simulations and projections under IPCC AR4. Theor. Appl. Climatol. 2007, 90, 133–159. [Google Scholar] [CrossRef]
  16. Mishra, A.K.; Dubey, A.K.; Dinesh, A.S. Diagnosing whether the increasing horizontal resolution of regional climate model inevitably capable of adding value: Investigation for Indian summer monsoon. Clim. Dyn. 2023, 60, 1925–1945. [Google Scholar] [CrossRef]
  17. Mishra, A.K.; Jangir, B.; Strobach, E. Does increasing climate model horizontal resolution be beneficial for the Mediterranean region?: Multimodel evaluation framework for High-Resolution Model Intercomparison Project. J. Geophys. Res. Atmos. 2023, 128, e2022JD037812. [Google Scholar] [CrossRef]
  18. Kim, G.; Kim, J.; Cha, D.H. Added value of high-resolution regional climate model in simulating precipitation based on the changes in kinetic energy. Geosci. Lett. 2022, 9, 38. [Google Scholar] [CrossRef]
  19. Devanand, A.; Ghosh, S.; Paul, S.; Karmakar, S.; Niyogi, D. Multi-ensemble regional simulation of Indian monsoon during contrasting rainfall years: Role of convective schemes and nested domain. Clim. Dyn. 2018, 50, 4127–4147. [Google Scholar] [CrossRef]
  20. Halenka, T.; Kalvová, J.; Chládová, Z.; Demeterová, A.; Zemánková, K.; Belda, M. On the capability of RegCM to capture extremes in long term regional climate simulation-comparison with the observations for Czech Republic. Theor. Appl. Clim. 2006, 86, 125–145. [Google Scholar] [CrossRef]
  21. Herrera, S.; Fita, L.; Fernández, J.; Gutiérrez, J.M. Evaluation of the mean and extreme precipitation regimes from the ENSEMBLES regional climate multimodel simulations over Spain. J. Geophys. Res. Atmos. 2010, 115, 21117. [Google Scholar] [CrossRef]
  22. Karmacharya, J.; Jones, R.; Moufouma-Okia, W.; New, M. Evaluation of the added value of a high-resolution regional climate model simulation of the South Asian summer monsoon climatology. Int. J. Climatol. 2017, 37, 3630–3643. [Google Scholar] [CrossRef]
  23. Karmacharya, J.; New, M.; Jones, R.; Levine, R. Added value of a high-resolution regional climate model in simulation of intraseasonal variability of the South Asian summer monsoon. Int. J. Climatol. 2017, 37, 1100–1116. [Google Scholar] [CrossRef]
  24. Maity, S.; Satyanarayana, A.N.V.; Mandal, M.; Nayak, S. Performance evaluation of land surface models and cumulus convection schemes in the simulation of Indian summer monsoon using a regional climate model. Atmos. Res. 2017, 197, 21–41. [Google Scholar] [CrossRef]
  25. Wang, Y.; Sen, O.L.; Wang, B. A Highly Resolved Regional Climate Model (IPRC-RegCM) and Its Simulation of the 1998 Severe Precipitation Event over China. Part I: Model Description and Verification of Simulation. J. Clim. 2003, 16, 1721–1738. [Google Scholar] [CrossRef]
  26. Steiner, A.L.; Pal, J.S.; Rauscher, S.A.; Bell, J.L.; Diffenbaugh, N.S.; Boone, A.; Sloan, L.C.; Giorgi, F. Land surface coupling in regional climate simulations of the West African monsoon. Clim. Dyn. 2009, 33, 869–892. [Google Scholar] [CrossRef]
  27. Mishra, A.K.; Dubey, A.K. Sensitivity of convective parameterization schemes in regional climate model: Precipitation extremes over India. Theor. Appl. Climatol. 2021, 146, 293–309. [Google Scholar] [CrossRef]
  28. Dash, S.K.; Pattnayak, K.C.; Panda, S.K.; Vaddi, D.; Mamgain, A. Impact of domain size on the simulation of Indian summer monsoon in RegCM4 using mixed convection scheme and driven by HadGEM2: Impact of domain size on ISM simulations. Clim. Dyn. 2014, 44, 961–975. [Google Scholar] [CrossRef]
  29. Halder, S.; Dirmeyer, P.A.; Saha, S.K. Sensitivity of the mean and variability of Indian summer monsoon to land surface schemes in RegCM4: Understanding coupled land-atmosphere feedbacks. J. Geophys. Res. Atmos. 2015, 120, 9437–9458. [Google Scholar] [CrossRef]
  30. Maurya, R.K.S.; Sinha, P.; Mohanty, M.R.; Mohanty, U.C. RegCM4 model sensitivity to horizontal resolution and domain size in simulating the Indian summer monsoon. Atmos. Res. 2018, 210, 15–33. [Google Scholar] [CrossRef]
  31. Ghosh, S.; Bhatla, R.; Mall, R.K.; Srivastava, P.K.; Sahai, A.K. Aspect of ECMWF downscaled Regional Climate Modeling in simulating Indian summer monsoon rainfall and dependencies on lateral boundary conditions. Theor. Appl. Climatol. 2019, 135, 1559–1581. [Google Scholar] [CrossRef]
  32. Mishra, A.K.; Dwivedi, S. Assessment of convective parametrization schemes over the Indian subcontinent using a regional climate model. Theor. Appl. Climatol. 2019, 137, 1747–1764. [Google Scholar] [CrossRef]
  33. Ajay, P.; Pathak, B.; Solmon, F.; Bhuyan, P.K.; Giorgi, F. Obtaining best parameterization scheme of RegCM 4.4 for aerosols and chemistry simulations over the CORDEX South Asia. Clim. Dyn. 2019, 53, 329–352. [Google Scholar] [CrossRef]
  34. Maurya, R.K.S.; Sinha, P.; Mohanty, M.R.; Mohanty, U.C. Coupling of community land model with RegCM4 for Indian summer monsoon simulation. Pure Appl. Geophys. 2017, 174, 4251–4270. [Google Scholar] [CrossRef]
  35. Dash, S.K.; Mishra, S.K.; Pattnayak, K.C.; Mamgain, A.; Mariotti, L.; Coppola, E.; Giorgi, F.; Giuliani, G.; Dash, S.K.; Mishra, S.K.; et al. Projected seasonal mean summer monsoon over India and adjoining regions for the twenty-first century. Theor. Appl. Climatol. 2015, 122, 581–593. [Google Scholar] [CrossRef]
  36. Jones, R.G.; Murphy, J.M.; Noguer, M. Simulation of climate change over Europe using a nested regional-climate model. I: Assessment of control climate, including sensitivity to location of lateral boundaries. Q. J. R. Meteorol. Soc. 1995, 121, 1413–1449. [Google Scholar]
  37. Seth, A.; Giorgi, F. The Effects of Domain Choice on Summer Precipitation Simulation and Sensitivity in a Regional Climate. J. Clim. 1998, 11, 2698–2712. Available online: https://journals.ametsoc.org/view/journals/clim/11/10/1520-0442_1998_011_2698_teodco_2.0.co_2.xml (accessed on 15 July 2022). [CrossRef]
  38. Elguindi, N.; Bi, X.; Giorgi, F.; Nagarajan, B.; Pal, J.; Solmon, F.; Rauscher, S.; Zakey, A.; O’Brien, T.; Nogherotto, R.; et al. Regional Climate Model RegCM Reference Manual; ITCP: Trieste, Italy, 2014. [Google Scholar]
  39. Dickinson, R.E.; Henderson-Sellers, A.; Kennedy, P.J. Biosphere-atmosphere Transfer Scheme (BATS) Version 1e as Coupled to the NCAR Community Climate Model. In NCAR Technical Note NCAR/TN-387+STR; National Center for Atmospheric Research: Boulder, CO, USA, 1993; p. 72. [Google Scholar] [CrossRef]
  40. Oleson, K.W.; Lead, D.M.L.; Bonan, G.B.; Drewniak, B.; Huang, M.; Koven, C.D.; Levis, S.; Li, F.; Riley, W.J.; Subin, Z.M.; et al. Technical description of version 4.5 of the Community Land Model (CLM). NCAR 2013, 257, 1–257. [Google Scholar] [CrossRef]
  41. Dee, D.P.; Uppala, S.M.; Simmons, A.J.; Berrisford, P.; Poli, P.; Kobayashi, S.; Andrae, U.; Balmaseda, M.A.; Balsamo, G.; Bauer, P.; et al. The ERA-Interim reanalysis: Configuration and performance of the data assimilation system. Q. J. R. Meteorol. Soc. 2011, 137, 553–597. [Google Scholar] [CrossRef]
  42. Reynolds, R.W.; Rayner, N.A.; Smith, T.M.; Stokes, D.C.; Wang, W. An Improved In Situ and Satellite SST Analysis for Climate. J. Clim. 2002, 15, 1609–1625. Available online: https://journals.ametsoc.org/view/journals/clim/15/13/1520-0442_2002_015_1609_aiisas_2.0.co_2.xml (accessed on 15 July 2022). [CrossRef]
  43. Pai, D.S.; Sridhar, L.; Rajeevan, M.; Sreejith, O.P.; Satbhai, N.S.; Mukhopadhyay, B. Development of a new high spatial resolution (0.25° × 0.25°) long period (1901–2010) daily gridded rainfall data set over India and its comparison with existing data sets over the region. Mausam 2014, 65, 1–18. [Google Scholar] [CrossRef]
  44. Hersbach, H.; Bell, B.; Berrisford, P.; Hirahara, S.; Horányi, A.; Muñoz-Sabater, J.; Nicolas, J.; Peubey, C.; Radu, R.; Schepers, D.; et al. The ERA5 global reanalysis. Q. J. R. Meteorol. Soc. 2020, 146, 1999–2049. [Google Scholar] [CrossRef]
  45. Giorgi, F.; Jones, C.; Asrar, G.R. Addressing Climate Information Needs at the Regional Level: The CORDEX Framework. World Meteorol. Organ. (WMO) Bull. 2009, 58, 175. Available online: https://public.wmo.int/en/bulletin/addressing-climate-information-needs-regional-level-cordex-framework (accessed on 15 July 2022).
  46. Kumari, A.; Kumar, P.; Dubey, A.K.; Mishra, A.K.; Saharwardi, M.S. Dynamical and thermodynamical aspects of precipitation events over India. Int. J. Climatol. 2022, 42, 3094–3106. [Google Scholar] [CrossRef]
  47. Singh, S.; Mall, R.K.; Singh, N. Changing spatio-temporal trends of heat wave and severe heat wave events over India: An emerging health hazard. Int. J. Climatol. 2021, 41, E1831–E1845. [Google Scholar] [CrossRef]
  48. Torma, C.; Giorgi, F.; Coppola, E. Added value of regional climate modeling over areas characterized by complex terrain—Precipitation over the Alps. J. Geophys. Res. Atmos. 2015, 120, 3957–3972. [Google Scholar] [CrossRef]
  49. Shahi, N.K.; Das, S.; Ghosh, S.; Maharana, P.; Rai, S. Projected changes in the mean and intra-seasonal variability of the Indian summer monsoon in the RegCM CORDEX-CORE simulations under higher warming conditions. Clim. Dyn. 2021, 57, 1489–1506. [Google Scholar] [CrossRef]
  50. Bhate, J.; Kesarkar, A.; Rajasree, V.P.M. Simulation of the diurnal cycle of rainfall during Indian summer monsoon season using mesoscale model. Theor. Appl. Climatol. 2019, 138, 185–200. [Google Scholar] [CrossRef]
  51. Fildier, B.; Parishani, H.; Collins, W.D. Simultaneous characterization of mesoscale and convective-scale tropical rainfall extremes and their dynamical and thermodynamic modes of change. J. Adv. Model. Earth Syst. 2017, 9, 2103–2119. [Google Scholar] [CrossRef]
  52. Flynn, W.J.; Nesbitt, S.W.; Anders, A.M.; Garg, P. Mesoscale precipitation characteristics near the Western Ghats during the Indian Summer Monsoon as simulated by a high-resolution regional model. Q. J. R. Meteorol. Soc. 2017, 143, 3070–3084. [Google Scholar] [CrossRef]
  53. Roca, R.; Fiolleau, T. Extreme precipitation in the tropics is closely associated with long-lived convective systems. Commun. Earth Environ. 2020, 1, 18. [Google Scholar] [CrossRef]
  54. Sillmann, J.; Thorarinsdottir, T.; Keenlyside, N.; Schaller, N.; Alexander, L.V.; Hegerl, G.; Seneviratne, S.I.; Vautard, R.; Zhang, X.; Zwiers, F.W. Understanding, modeling and predicting weather and climate extremes: Challenges and opportunities. Weather Clim. Extrem. 2017, 18, 65–74. [Google Scholar] [CrossRef]
  55. Mishra, S.K.; Sahany, S.; Salunke, P.; Kang, I.S.; Jain, S. Fidelity of CMIP5 multi-model mean in assessing Indian monsoon simulations. NPJ Clim. Atmos. Sci. 2018, 1, 39. [Google Scholar] [CrossRef]
  56. Tiwari, P.R.; Kar, S.C.; Mohanty, U.C.; Dey, S.; Sinha, P.; Raju, P.V.; Shekhar, M.S. The role of land surface schemes in the regional climate model (RegCM) for seasonal scale simulations over Western Himalaya. Atmósfera 2015, 28, 129–142. [Google Scholar] [CrossRef]
  57. Zheng, Y.Q.; Qian, Y.F.; Miao, M.Q.; Yu, G.; Kong, Y.; Zhang, D. The effects of vegetation change on regional climate I: Simulation results. Acta Meteorol. Sin. 2002, 60, 1–16. [Google Scholar]
Atmosphere 15 00044 g001
Figure 1. The upper panel represents the model domains and topography of the region. The outer domain represents the bigger South Asia Coordinated Regional Climate Downscaling Experiment (CORDEX) domain (SACD). The white box in the inset represents the smaller Indian (IND) domain. The subregions inside India represent the homogeneous rainfall regions. The shaded color map represents the topography of the region. The lower panel is the graphical representation of the methodology adopted.
Figure 1. The upper panel represents the model domains and topography of the region. The outer domain represents the bigger South Asia Coordinated Regional Climate Downscaling Experiment (CORDEX) domain (SACD). The white box in the inset represents the smaller Indian (IND) domain. The subregions inside India represent the homogeneous rainfall regions. The shaded color map represents the topography of the region. The lower panel is the graphical representation of the methodology adopted.
Atmosphere 15 00044 g001
Atmosphere 15 00044 g002
Figure 2. JJAS mean R99 (upper panel) and R95 (lower panel) for all simulations and IMD during 2006–2015.
Figure 2. JJAS mean R99 (upper panel) and R95 (lower panel) for all simulations and IMD during 2006–2015.
Atmosphere 15 00044 g002
Atmosphere 15 00044 g003
Figure 3. JJAS mean R10 (upper panel) and R50 (lower panel) for all simulations and IMD during 2006–2015.
Figure 3. JJAS mean R10 (upper panel) and R50 (lower panel) for all simulations and IMD during 2006–2015.
Atmosphere 15 00044 g003
Atmosphere 15 00044 g004
Figure 4. Percentage bias (IMD-model) for R95 (upper panel (a–d)) and R99 (lower panel (e–h)) from 2006 to 2015.
Figure 4. Percentage bias (IMD-model) for R95 (upper panel (a–d)) and R99 (lower panel (e–h)) from 2006 to 2015.
Atmosphere 15 00044 g004
Atmosphere 15 00044 g005
Figure 5. Percentage bias (IMD model) for R10 (upper panel (a–d)) and R50 (lower panel (e–h)) from 2006 to 2015.
Figure 5. Percentage bias (IMD model) for R10 (upper panel (a–d)) and R50 (lower panel (e–h)) from 2006 to 2015.
Atmosphere 15 00044 g005
Atmosphere 15 00044 g006
Figure 6. Daily precipitation intensity-probability distribution functions (PDFs) (frequency versus intensity of daily precipitation events) for all simulation and IMD during JJAS of 2006–2015 for different Indian homogeneous regions, (a) Central North India, (b) West Central India, (c) Northeast India (d) Southern Peninsular India, (e) Hilly Region, (f) Northwest India. The unit of precipitation intensity is mm, and the frequency is days−1.
Figure 6. Daily precipitation intensity-probability distribution functions (PDFs) (frequency versus intensity of daily precipitation events) for all simulation and IMD during JJAS of 2006–2015 for different Indian homogeneous regions, (a) Central North India, (b) West Central India, (c) Northeast India (d) Southern Peninsular India, (e) Hilly Region, (f) Northwest India. The unit of precipitation intensity is mm, and the frequency is days−1.
Atmosphere 15 00044 g006
Atmosphere 15 00044 g007
Figure 7. Consecutive dry days (lower panel (e–h)) and consecutive wet days (upper panel (a–d)) (in days/year) for all simulations and IMD for JJAS from 2006 to 2015.
Figure 7. Consecutive dry days (lower panel (e–h)) and consecutive wet days (upper panel (a–d)) (in days/year) for all simulations and IMD for JJAS from 2006 to 2015.
Atmosphere 15 00044 g007
Atmosphere 15 00044 g008
Figure 8. Composite of convective precipitation (left panel) and large-scale precipitation (right panel) during the extreme precipitation days for ERA5.
Figure 8. Composite of convective precipitation (left panel) and large-scale precipitation (right panel) during the extreme precipitation days for ERA5.
Atmosphere 15 00044 g008
Atmosphere 15 00044 g009
Figure 9. The composite of convective scale precipitation for all simulations for the extreme precipitation days.
Figure 9. The composite of convective scale precipitation for all simulations for the extreme precipitation days.
Atmosphere 15 00044 g009
Atmosphere 15 00044 g010
Figure 10. The composite of large-scale precipitation for all simulations for the extreme precipitation days.
Figure 10. The composite of large-scale precipitation for all simulations for the extreme precipitation days.
Atmosphere 15 00044 g010
Atmosphere 15 00044 g011
Figure 11. The composite of low-level (850 hPa) circulation (vector) and specific humidity (shaded) for all simulations and corresponding observation for the extreme precipitation days.
Figure 11. The composite of low-level (850 hPa) circulation (vector) and specific humidity (shaded) for all simulations and corresponding observation for the extreme precipitation days.
Atmosphere 15 00044 g011
Atmosphere 15 00044 g012
Figure 12. The composite of atmospheric stability for all simulations for the extreme precipitation days.
Figure 12. The composite of atmospheric stability for all simulations for the extreme precipitation days.
Atmosphere 15 00044 g012
Table 1. Description of the experiments, model configuration, and physics.
Table 1. Description of the experiments, model configuration, and physics.
Land surface modelBATSCLM4.5
DomainSACDINDSACDIND
Experiment nameBATSSABATSINDCLMSACLMIND
Soil temperatures calculationUses a two-layer force–restore modelSoil temperature is calculated explicitly by a 10-layer soil model
Surface representationOne vegetation layer, a surface soil layer, a snow layerOne vegetation layer with a canopy photosynthesize conductance model, 10 unevenly spaced soil layers, five snow layers with an additional representation of trace snow
Treatment of heat and roughness lengthHeat and water vapor roughness lengths are constantUpdates these values over bare soil and snow with values from the stability functions
Albedo treatmentUses prescribed values for vegetation albedo for both short- and longwave componentsUses a modified two-stream approach that reduces the complexity of a full two-stream albedo treatment
Treatment of heat and roughness lengthHeat and water vapor roughness lengths are constantUpdates these values over bare soil and snow with values from the stability functions
Land cover/vegetation classes2024
Treatment of vegetation canopyTreats all vegetation within the canopy in the same manner The canopy is divided into sunlit and shaded fractions as a function of LAI
Calculation of stomatal conductance and photosynthesis rateNo individual calculation is made for sunlit and shaded fractions. It does not compute photosynthetic ratesStomatal conductance is calculated for sunlit and shaded fractions. Calculation of photosynthetic rates is made in this scheme
Cumulus parameterizationMIT over land and Tiedke over the ocean
PBL parameterizationUW PBL scheme
Radiation parameterizationCCM3
Horizontal resolution50 km
Vertical layer23 having terrain-following Sigma coordinate
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Mishra, A.K.; Dinesh, A.S.; Kumari, A.; Pandey, L.K. Precipitation Extremes over India in a Coupled Land–Atmosphere Regional Climate Model: Influence of the Land Surface Model and Domain Extent. Atmosphere 2024, 15, 44. https://doi.org/10.3390/atmos15010044

AMA Style

Mishra AK, Dinesh AS, Kumari A, Pandey LK. Precipitation Extremes over India in a Coupled Land–Atmosphere Regional Climate Model: Influence of the Land Surface Model and Domain Extent. Atmosphere. 2024; 15(1):44. https://doi.org/10.3390/atmos15010044

Chicago/Turabian Style

Mishra, Alok Kumar, Anand Singh Dinesh, Amita Kumari, and Lokesh Kumar Pandey. 2024. "Precipitation Extremes over India in a Coupled Land–Atmosphere Regional Climate Model: Influence of the Land Surface Model and Domain Extent" Atmosphere 15, no. 1: 44. https://doi.org/10.3390/atmos15010044

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop