Quantification of agricultural drought over Indian region: a multivariate phenology-based approach

Abstract

The objective, accurate and rapid quantification of agricultural drought is the key component of effective drought planning and management mechanism. The present study proposed a new index, i.e. multivariate phenology-based agricultural drought index (MADI), for quantification of the agricultural drought using long-term (1982–2015) crop phenological parameters. The 15-day global inventory modelling and mapping studies time-series normalized difference vegetation index (NDVI) data (~ 8 km) were interpolated at daily scale and smoothened using Savitzky and Golay filtering technique. Different crop phenological parameters, i.e. start of season, end of season, length of the growing period (lgp), integrated NDVI (iNDVI), etc., were estimated using a combination of threshold and derivative approaches for individual pixels during kharif season. Based on the time of occurrence, the agricultural droughts may lead to delay in crop sowing, reduction in cropped area and/or decreased production. Hence, the lgp and iNDVI were selected among all phenological parameters for their capability to represent alterations in crop duration and crop production, respectively. The long-term lgp and iNDVI of individual pixel were detrended and transformed into standardized lgp (Slgp) and standardized iNDVI (SiNDVI) to eliminate the existing trends developed due to technological improvements during study period and existing heterogeneity of Indian agricultural system, respectively. The MADI was calculated by fitting Slgp and SiNDVI into joint probability distribution, where the best joint distribution family along with associated parameters was selected based on the goodness-of-fit for individual pixel. The values of MADI vary between − 4 and + 4, where the negative and positive values represent drought and non-drought conditions, respectively. The efficacy of the proposed index was tested over the Indian region by comparing with the multivariate standardized drought index, which considers the impacts of both meteorological and soil moisture drought using copula approach.