A Comparative Study of Four Merging Approaches for Regional Precipitation Estimation

Zedong Fan,Weiwei Sun,Jiahong Wen,Jun Gao,Weiyue Li,Qin Jiang

doi:10.1109/access.2021.3057057

Abstract

To identify suitable merging methods to improve regional precipitation estimates using multiple sources of precipitation data, this study applied four different approaches (multiple linear regression (MLR), feedforward neural network (FNN), random forest (RF) and long short-term memory network (LSTM)) to merge four satellite precipitation products and one reanalysis data in the Jiangsu, Zhejiang and Shanghai of China. The pros and cons of the merging approaches are analyzed comprehensively, using correlation coefficient (CC), root mean square error (RMSE), relative bias (RB), probability of detection (POD), false alarm ratio (FAR), and critical success index (CSI) as evaluation indexes. Our results show that: (1) All merging approaches can improve the accuracy of precipitation estimations, but only RF and LSTM can improve the daily precipitation event detection capacity. These approaches can significantly reduce errors in moderate precipitation scenarios, but do not effectively improve accuracy in light and heavy precipitation scenarios. (2) MLR was the least expensive computing cost method in our study and performed better than the other three methods when gauge density was low. However, MLR had the worst daily precipitation event detection capacity (CSI = 0.67). (3) FNN performed moderately in most experiments (CC = 0.87, RMSE = 4.65 mm/day, RB = 1.19 %, POD = 0.94, FAR = 0.29, CSI = 0.70). (4) The merged data generated by RF was the most accurate and had the best daily precipitation event detection capacity (CC = 0.87, RMSE = 4.61 mm/day, RB = − 0.33 %, POD = 0.97, FAR = 0.20, CSI = 0.78). RF performed best in moderate precipitation scenarios. However, it performed worse than other methods when gauge density was low. (5) LSTM was the most robust methods and performed best in light precipitation scenarios. The FAR of the LSTM-generated data was the smallest (0.15) among four fusion methods. However, LSTM had the most expensive computing cost and the worst accuracy of the merged data (CC = 0.86, RMSE = 4.68 mm/day, RB = − 9.36 %).

Highlights

Quantitative and accurate precipitation estimation is crucial for water resource management, natural disaster prevention, and risk management [1], [2]
(4) The merged data generated by random forest (RF) was the most accurate and had the best daily precipitation event detection capacity (CC = 0.87, root mean square error (RMSE) = 4.61 mm/day, relative bias (RB) = − 0.33 %, probability of detection (POD) = 0.97, false alarm ratio (FAR) = 0.20, critical success index (CSI) = 0.78)
M-multiple linear regression (MLR), for CHIRPS, ERA5, and the four merged datasets in the M-feedforward neural networks (FNN) and M-RF all performed significantly better than the precipitation products, with more than 85% of pixels having RBs between − 10% and 10%

Summary

Introduction

Quantitative and accurate precipitation estimation is crucial for water resource management, natural disaster prevention, and risk management [1], [2]. Precipitation products, satellite precipitation products and reanalysis datasets) have their own advantages and disadvantages. Z. Fan et al.: Comparative Study of Four Merging Approaches for Regional Precipitation Estimation and reanalysis datasets have the advantages of quasi global spatial coverage. Satellite rainfall products can provide near real-time data [5], and reanalysis datasets have a long history of data records [6], which have been widely used in climate change research [7]. Retrieval algorithms, and spatial sampling frequencies, can often cause satellite precipitation products to present regionally varying biases [9]. Due to the influences of the assimilation model [10] and observation data error [8], there can be spatial differences in precipitation estimation accuracy [11]. Many scholars have attempted to obtain more accurate precipitation data by correcting these raw precipitation products or even by combining multiple products

Methods

Results

Discussion

Conclusion