Comparative Analysis of Drought Based on Precipitation and Soil Moisture Indices in Haihe Basin of North China During the Period of 1960–2010

Yue Qin,Dawen Yang,Huimin Lei,Kai Xu,Xiangyu Xu
DOI: https://doi.org/10.1016/j.jhydrol.2014.09.068
IF: 6.4
2015-01-01
Journal of Hydrology
Abstract:Drought severity not only depends on weather anomaly, but is also related to terrestrial hydrological condition to a large extent. In this study, we analyzed droughts using indices based on precipitation and soil moisture during the period of 1960-2010 in Haihe basin, which is a typical drought-prone region in North China. The Soil Moisture Drought Severity (SMDS) and Standardized Precipitation Index (SPI) are used to evaluate drought severity. SMDS is calculated based on the monthly soil moisture of upper 50 cm from the simulation by Community Land Model (CLM 4.0) and SPI is calculated based on gridded precipitation at 0.05 degrees resolution (5 km x 5 km approximately), which is spatially interpolated from observations. During the last 51 years, 36 severe drought events (affecting areas greater than 20,000 km(2) and durations longer than 3 months) have been identified based on SMDS, and 41 drought events identified based on SPI. Results derived from SMDS indicate that there is a significant increasing trend in the drought affected area, and that the drought event occurred in 1999 has the largest affected area. Compared with the drought events derived from SMDS, the events derived from SPI have shorter durations but larger affected areas on average. Although the mean NDVI of the whole basin has been increasing since the 1980s, the two declining periods of 1992-1994 and 1999-2003 show fairly good agreement with the drought events identified in the same periods. The Anomaly of Normalized Difference Vegetation Index (A-NDVI) is introduced as NDVI anomaly from its trend line, thus the negative value of A-NDVI can reflect the drought impact on vegetation reasonably. Result indicates that both the SMDS and SPI are significantly correlated with A-NDVI, and correlation between annual SMDS and A-NDVI is higher than that of SPI. (C) 2014 Elsevier B.V. All rights reserved.
What problem does this paper attempt to address?