High-resolution one-day probable maximum precipitation dataset across India and its future-projected changes over India.

This article presents a high-resolution (0.25° latitude x 0.25° longitude) one-day Probable Maximum Precipitation (PMP) estimates across India based on the daily precipitation records after climate regime shift in 1970s (1971–2010). In addition to this, possible spatio-temporal changes in the PMP estimates in future towards the end of this century (2071–2100) with respect to its current values are also presented, following two possible emission scenarios (RCP4.5 and RCP8.5). These maps are developed following the research article entitled “Increase in Probable Maximum Precipitation in a Changing Climate over India”, published in Journal of Hydrology by Sarkar and Maity [1] .The recent PMP map for India is developed based on gridded daily observational records of precipitation, procured from the India Meteorological Department (IMD). Its future projection is based on bias-corrected simulated daily precipitation output from three Regional Climate Models (RCMs). Finally, the change in PMP in future with respect to the current map is calculated in terms of grid-wise percentage change and made available to the end-users. This PMP map and its future projection will serve as an important information for the design engineers and hydro-meteorologists for planning and designing various water-energy infrastructures such as, dams and other major water resources engineering applications.


a b s t r a c t
This article presents a high-resolution (0.25 °latitude x 0.25 °longitude) one-day Probable Maximum Precipitation (PMP) estimates across India based on the daily precipitation records after climate regime shift in 1970s . In addition to this, possible spatio-temporal changes in the PMP estimates in future towards the end of this century (2071-2100) with respect to its current values are also presented, following two possible emission scenarios (RCP4.5 and RCP8.5). These maps are developed following the research article entitled "Increase in Probable Maximum Precipitation in a Changing Climate over India", published in Journal of Hydrology by Sarkar and Maity [1] .The recent PMP map for India is developed based on gridded daily observational records of precipitation, procured from the India Meteorological Department (IMD). Its future projection is based on bias-corrected simulated daily precipitation output from three Regional Climate Models (RCMs). Finally, the change in PMP in future with respect to the current map is calculated in terms of grid-wise percentage change and made available to the end-users. This PMP map and its future projection will serve as an important information for the design engineers and hydro-meteorologists for planning and designing various water-energy infrastructures such as, dams and other major water resources engineering applications. The future-simulated precipitation data (2011-2100) was obtained for the domain named "WAS-44". Then from this domain, data for Indian mainland was extracted and bias-corrected. Description of data collection • The gridded precipitation data set was prepared by India Meteorological Department (IMD) based on the daily rainfall records at 6955 rain gauge stations across India from 1901-2010 [3] . This data was procured from the IMD to filter out the grid wise AMDP series and to compute PMP estimates for the post-1970 period.
• Simulated precipitation output from three Global Circulation Models (GCMs), downscaled by different Regional Climate models (RCMs) following two possible emission scenarios (RCP 4.5 and RCP 8.5) are downloaded from the Coordinated Regional Downscaling Experiment (CORDEX) portal: https://esgf-data.dkrz.de/search/cordex-dkrz/ . These are bias-corrected and AMDP values are extracted to estimate PMP using data towards the end of the century (2071-2100 Value of the data • Recent research findings established a significant increase in PMP. Thus, the PMP estimates with the data after climate regime shift (post-1970s) must be considered for design of various water infrastructure. • Spatio-temporal change in future-projected PMP over India provides a meaningful insight to the consequence of climate change. This information must be considered for the design of the hydraulic structures with long return periods. • Design engineers and policy makers can use these data to revise the design of existing infrastructures due to increased risk in the context of climate change with the passage of time. • Additionally, the future bias-corrected Annual Maximum Daily Precipitation (AMDP) series can be used extensively in different studies on future possible changes in precipitation extremes over India.

Data description
This dataset contains the gridded AMDP series and PMP estimates over India, based on recent IMD observations . Additionally, it also represents multi-model AMDP series based on bias-corrected future-simulated precipitation outputs from three models (M1, M2, and M3), following two emission scenarios (RCP 4.5 and RCP 8.5) for the far-future period 2071-2100. The corresponding model-averaged (across M1, M2 and M3) changes in PMP over India w.r.t. the current values (1971-2010) are shown in Fig. 1 and Fig. 2    Similarly, the folder 'Based on future-simulated data (2071-2100)' contains ten files; eight NETCDF files and two TIFF files, described as follows

Experimental design, materials, and methods
The observational dataset of PMP over the entire Indian mainland is determined based on daily gridded (0.25 °latitude × 0.25 °longitude) rainfall data, procuring from India Meteorolog-ical Department (IMD). The length of the data is 110 years, i.e., from 1901to 2010. However, in order to develop the recent PMP estimates, data for the last 40 years (1971-2010) is used.
To determine the possible changes in PMP towards the end of this century, daily simulated precipitation values for the future time period 2011-2100 (i.e., 90 years) is obtained from the Coordinated Regional Downscaling Experiment (CORDEX) portal, a reliable data archive, supported by World Climate Research Programme (WCRP). The future-simulated precipitation values are obtained for three different Regional Climate Models (RCMs), for two emission scenarios, designated by Representative Concentration Pathways (RCP 4.5 and RCP 8.5) from the CORDEX database. For further details on the data, including its driving model, source institute, data resolution for those three different model-combinations (named as M1, M2, and M3), readers can refer to Table-1 of the companion 'Journal of Hydrology' article [1] . To study the change in PMP towards the end of this century with respect to the current estimates (1971-2010), simulated future precipitation values are used during 2071-2100.
Future-simulated precipitation values are bias-corrected before further analysis. A recently developed copula-based bias-correction technique by Maity et al. [4] is applied, which is proven to correct the bias in both mean and extreme values, and also suitable for zero-inflated daily simulated precipitation series.
In case of both observed and future daily precipitation data, first the Annual Maximum Daily Precipitation (AMDP) time series is extracted, which is stored in the data repository. Then the method proposed by Sarkar and Maity [ 1 , 2 ] is applied on the AMDP series to estimate PMP. Finally, the recent PMP estimates and maps are developed and uploaded in the data repository. Similarly, multi-model averaged (average of M1, M2, and M3) PMP estimate towards the end of the 21st century are developed and changes are captured w.r.t. the current PMP estimates. The changes in terms of percentage difference are shown in the form of maps and also stored in the data-repository in NetCDF format.