Evaluation of precipitation datasets available on Google earth engine over India

Abstract

Monthly mean precipitation estimates of seven products (TerraClimate, TRMM, CHIRPS, PERSIANN‐CDR, GPM‐IMERG, ERA5 and CFSR) available on Google earth engine (GEE) are evaluated against gridded gauge‐based precipitation product available from Indian Meteorological Department (IMD) for their skills and presence of systematic biases (during 2001–2018). All these products represent the climatological features reasonably well. Presence of systematic biases in these products is also observed from their evaluation. Biases across the periphery of the country are relatively on the higher side in comparison to the central regions. The magnitude of spatial variability is represented better for winter precipitation in comparison to summer precipitation. During both winter and summer, ensemble mean of various products outperforms individual products in terms of both RMSE and correlation. Performance of these products is also assessed across various Indian states, elevation bands and climate zones. The ability of these products to represent the seasonality was observed to be highest for the states with mid‐ranged peaks (10–20 mm·day−1) which tend to decrease with both increasing and decreasing peaks. Ability of the precipitation products to resemble the annual cycle does not vary with the amount of precipitation, although individual disparity among the products exists. Additionally, an alternative approach for data evaluation using Multiple Triple Collocation (MTC) was performed for the period 2001–2015 using an additional dataset obtained from soil‐moisture‐based rainfall estimates (SM2RAIN). Results from MTC convey that ERA5 performs relatively poor in comparison to the other products for central India followed by CFSR. In brief, the comprehensive evaluation of precipitation products reported herein will act a valuable reference for the researchers as well as decision makers to select the optimal product for their intended application and will inform the users about the various uncertainties in the foundations and specification of these products.

Evaluation of IMERG and ERA5 precipitation products over the Mongolian Plateau

“…Tang et al concluded that IMERG generally outperformed ERA5 across China, and can better reproduce precipitation diurnal cycles 55 . Other studies were presented in Central Asia 56 , India 57 , Turkey 47 , Iran 58 , and the United States 59 . Most studies have shown that IMERG outperforms ERA5, but the superiority of each dataset varies by regions, precipitation intensity, and altitude.…”

Streamflow Prediction in Highly Regulated, Transboundary Watersheds Using Multi‐Basin Modeling and Remote Sensing Imagery

“…We therefore evaluated five operational products using different statistical metrics (CC, RMSE, and relative error [RE]; Table S2 in Supporting Information S1). It was revealed that each product has different strengths and weaknesses (in terms of correlation, errors, and mean bias) as found in previous studies (Dinh et al, 2020;Dubey et al, 2021;Le et al, 2020). Accordingly, all five products were merged using weighted average spatial correlation for the 536 rain gauges in the region.…”

Modeling canopy water content in the assessment for rainfall induced surface and groundwater nitrate contamination index from cropland N-fertilizer in Bilate downstream

“…The missing rainfall point data is computed using CHIRPS daily observation integrating longitude and latitude of the selected gauged station points to Google earth engine. which is very effective on estimation of precipitation in daily, monthly and annual based estimation (Banerjee et al, 2020;Dubey et al, 2021;C. Y. Liu et al, 2020;Mab et al, 2019;Molla et al, 2022;Wiwoho et al, 2021).…”