Quantile-based Bayesian Model Averaging approach towards merging of precipitation products

Abstract Precipitation is a fundamental input for many hydrological and water management studies. Nowadays, a number of satellite precipitation products are easily accessible online at free of cost. Despite so, the utility of such products is still limited owing to their lack of accuracy in capturing the ground truth. To improve the reliability of the satellite precipitation products, we have developed a quantile based Bayesian model averaging (QBMA) approach to merge the satellite precipitation products. QBMA approach was compared with traditional methods, namely, simple model averaging and one outlier removed. We have considered three SPPs (TRMM, PERSIANN-CDR, CMORPH) for QBMA merging during the monsoon season over India's coastal Vamsadhara river basin. QBMA optimal weights were trained using 2001 to 2013 daily monsoon precipitation data and validated for 2014 to 2018. Results indicated that the bias-corrected QBMA outperformed the other methods. On monthly evaluation, it is observed that all the products perform better during July and September than that in June and August. The QBMA approaches do not have any significant improvement over the SMA approach in terms of POD. However, the bias-corrected QBMA products have lower FAR. The developed QBMA approach with bias-corrected inputs outperforms the IMERG product in terms of RMSE.