Tracking seasonal and monthly drought with GRACE-based terrestrial water storage assessments over major river basins in South India

K Satish Kumar; E Venkata Rathnam; Venkataramana Sridhar

doi:10.1016/j.scitotenv.2020.142994

Tracking seasonal and monthly drought with GRACE-based terrestrial water storage assessments over major river basins in South India

Sci Total Environ. 2021 Apr 1:763:142994. doi: 10.1016/j.scitotenv.2020.142994. Epub 2020 Oct 15.

Authors

K Satish Kumar¹, E Venkata Rathnam², Venkataramana Sridhar³

Affiliations

¹ Department of Civil Engineering, National Institute of Technology, Warangal, India. Electronic address: satish005@student.nitw.ac.in.
² Department of Civil Engineering, National Institute of Technology, Warangal, India. Electronic address: evr@nitw.ac.in.
³ Deparment of Biological Systems Engineering, Virginia Polytechnic Institute and State University, Blacksburg, VA 24061, USA. Electronic address: vsri@vt.edu.

PMID: 33129527
DOI: 10.1016/j.scitotenv.2020.142994

Abstract

Drought is a complex natural hazard that affects ecosystems and society in several ways and it is important to quantify drought at the river basin scale. Assessment of drought requires both hydrological observations and simulation models as the data are generally scarce. Therefore, we use remote sensing products to help understand drought conditions in four basins in South India. This study analysed the correlation among five drought indices for four seasons: gravity recovery and climate experiment - drought severity index (GRACE-DSI), standardized precipitation index (SPI), self-calibrated palmer drought severity index (sc_PDSI), standardized precipitation-evapotranspiration index (SPEI), and combined climatologic deviation index (CCDI) with GRACE terrestrial water storage anomalies (TWSA) using the Pearson correlation coefficient (r) from 2002 to 2016 over the Godavari, Krishna, Pennar, and Cauvery river basins. Basin scale drought events are evaluated using CCDI, GRACEDSI, sc_PDSI, SPI12, and SPEI12 at seasonal and monthly time scale. Characteristics of drought event analysis are calculated for CCDI monthly. The results showed that GRACE TWS is highly correlated with GRACE-DSI, CCDI, and sc_PDSI. Seasonally, high spatial correlations between CCDI and GRACE-DSI with GRACE TWS are evident for all the river basins. Additionally, correlation is found to exist between sc_PDSI and GRACE TWS as soil moisture content is an operating variable between them. The 12-month SPI and SPEI correlated better with GRACE TWS than the 3 and 6-month periods. Among the four basins, droughts in the Krishna Basin lasted 29 months, longer than in the rest of the basins between 2003 and 2005. Overall, CCDI and GRACE-DSI indices are found to be effective for examining and evaluating the drought conditions at the basin scale.

Keywords: Drought characteristics; Drought indices; GRACE; Pearson correlation coefficient; TWSA.