Saudi Rainfall (SaRa): hourly 0.1° gridded rainfall (1979–present) for Saudi Arabia via machine learning fusion of satellite and model data

Wang, X; Alharbi, RS; Baez-Villanueva, OM; Green, A; McCabe, MF; Wada, Y; Van Dijk, AIJM; Abid, MA; Beck, HE

Journal article

Saudi Rainfall (SaRa): hourly 0.1° gridded rainfall (1979–present) for Saudi Arabia via machine learning fusion of satellite and model data

Abstract:: Abstract. We introduce Saudi Rainfall (SaRa), a gridded historical and near-real-time precipitation (P) product specifically designed for the Arabian Peninsula, one of the most arid, water-stressed, and data-sparse regions on Earth. The product has an hourly 0.1° resolution spanning 1979 to the present and is continuously updated with a latency of less than 2 h. The algorithm underpinning the product involves 18 machine learning model stacks trained for different combinations of satellite and (re)analysis P products along with several static predictors. As a training target, hourly and daily P observations from gauges in Saudi Arabia (n = 113) and globally (n = 14 256) are used. To evaluate the performance of SaRa, we carried out the most comprehensive evaluation of gridded P products in the region to date, using observations from independent gauges (randomly excluded from training) in Saudi Arabia as a reference (n = 119). Among the 20 evaluated P products, our new product, SaRa, consistently ranked first across all evaluation metrics, including the Kling–Gupta efficiency (KGE), correlation, bias, peak bias, wet-day bias, and critical success index. Notably, SaRa achieved a median KGE – a summary statistic combining correlation, bias, and variability – of 0.36, while widely used non-gauge-based products such as CHIRP, ERA5, GSMaP V8, and IMERG-L V07 achieved values of −0.07, 0.21, −0.13, and −0.39, respectively. SaRa also outperformed four gauge-based products such as CHIRPS V2, CPC Unified, IMERG-F V07, and MSWEP V2.8 which had median KGE values of 0.17, −0.03, 0.29, and 0.20, respectively. Our new P product – available at https://www.gloh2o.org/sara (last access: 24 September 2025) – addresses a crucial need in the Arabian Peninsula, providing a robust and reliable dataset to support hydrological modeling, water resource assessments, flood management, and climate research.

Publication status:: Published

Peer review status:: Peer reviewed

Actions

Email

Email this record

Send the bibliographic details of this record to your email address.

Your Email
Please enter the email address that the record information will be sent to.

-
Your message (optional)
Please add any additional information to be included within the email.
Share
Cite

Cite this record

APA Style

Wang, X., Alharbi, R. S., Baez-Villanueva, O. M., Green, A., McCabe, M. F., Wada, Y., Van Dijk, A. I. J. M., Abid, M. A., & Beck, H. E. (2025). Saudi Rainfall (SaRa): hourly 0.1° gridded rainfall (1979–present) for Saudi Arabia via machine learning fusion of satellite and model data. Hydrology and Earth System Sciences, 29(19), 4983–5003.

MLA Style

Wang, X, et al. “Saudi Rainfall (SaRa): Hourly 0.1° Gridded Rainfall (1979–Present) for Saudi Arabia via Machine Learning Fusion of Satellite and Model Data.” Hydrology and Earth System Sciences, vol. 29, no. 19, 2025, pp. 4983–5003.

Chicago Style

Wang, X, RS Alharbi, OM Baez-Villanueva, et al. 2025. “Saudi Rainfall (SaRa): Hourly 0.1° Gridded Rainfall (1979–Present) for Saudi Arabia via Machine Learning Fusion of Satellite and Model Data.” Hydrology and Earth System Sciences 29 (19): 4983–5003.
Print