Systematical Evaluation of Three Gridded Daily Precipitation Products Against Rain Gauge Observations Over Central Asia

Tuoliewubieke Dilinuer,Yong Zhao,Junqiang Yao,Lianmei Yang,Jiangang Li,Jing Chen,Weiyi Mao

doi:10.3389/feart.2021.699628

Abstract

Understanding the precipitation variability and extreme precipitation over arid Central Asia (CA) has largely been hampered by the lack of daily precipitation observations. The gridded precipitation datasets over CA are large discrepancies. Here, three gauge-based gridded daily precipitation products from Asian Precipitation Highly-Resolved Observational Data Integration Towards Evaluation (APHRODITE), Global Precipitation Climatology Center (GPCC), and Climate Prediction Center Based Analysis of Global Daily Precipitation (CPC_global) were assessed and compared with 49 rain gauge daily observations precipitation (OBS) from January 1985 to December 2015 using different time-scales over CA and different climate regimes, specifically Northern CA with temperate continental climate (NCA), Southwestern CA with dry arid desert climate (SWCA), and Southeastern CA with Mediterranean continental climate (SECA). Four accuracy indices [correlation coefficient (R), Bias, root mean square error (RMSE), and relative bias (RBias)] were employed to evaluate the performance of the three products in depicting the spatiotemporal features of precipitation variation over CA at multiple time scales (including daily, monthly, seasonal, and yearly). The mean annual and daily precipitation of OBS and three gridded products exhibit the trend of a gradual precipitation decreased from SECA to NCA and SWCA. The best overall performance was obtained for APHRODITE and GPCC for daily and annual time-scale, whereas CPC shows noticeable underestimation precipitation in SECA. The monthly precipitation depicted distinct features with a bimodal pattern with a peak in March and another in December, include the SECA and SWCA regions. In contrast, precipitation was concentrated in summer with the peak in July over the NCA region. At monthly scale terms, APHRODITE was more accurate in the wet seasons (winter and spring months) in SWCA and SECA. Additionally, GPCC has fairly better capability in summer months in NCA. Considering the spatial distribution, the bias variability was largerly in mountainous areas than in the plains. Temporally, the bias largerly in the dry seasons than in the wet seasons. At the interannual variability scale, GPCC was capable of qualitatively increasing the CA (NCA and SECA) precipitation during the last 21 years, while APHRODITE underestimated the trends. The CPC overestimated the precipitation trends over all regions. This study can serve as a reference for selecting daily precipitation products with low densities of stations, complex topographies, and similar climatic regions.

Highlights

Precipitation is an important component in hydrology and climate systems, and plays a vital role in driving land surfaceatmosphere interactions (Yao et al, 2019; Tan et al, 2020)
Mean annual precipitation exhibit the trend of a gradual precipitation decreased from southeastern to northwestern CA regions with precipitation hotspots located in Southeastern CA (SECA) and Northern Central Asia (NCA).The area with the lowest average annual precipitation expanded over the SWCA, which has an arid climate pattern and consists of desert areas (Figure 2A)
As for Bias (Figure 7I), the results show that APHRODITE underestimated the precipitation, except in July

Summary

Introduction

Precipitation is an important component in hydrology and climate systems, and plays a vital role in driving land surfaceatmosphere interactions (Yao et al, 2019; Tan et al, 2020). Due to sparseness and unevenness distribution of in-situ observation, or the discontinuity in recording sequences, they do not provide a reliable spatiotemporal representation of precipitation (Gruber and Levizzani, 2008) This is the case over complex regions where there is insufficient raingauge available and rainfall is distinguished by complex patterns especially over desert, mountainous terrain, and sparsely populated areas (Rana and McGregor, 2015; Yao et al, 2019). In this context, Gridded datasets are useful alternatives to ground observations and are widely used owing to their high resolution and open access (Serreze and Hurst, 2000; Silva et al, 2007; Pai et al, 2014; Guo et al, 2015a). The multiple gridded precipitation datasets have pros and cons, so it is necessary to validate the applicability of products before further utilization (Henn et al, 2018; Sun et al, 2018; Tan et al, 2020)

Methods

Results

Conclusion