The present study was conducted over Jharkhand state (India) for assessing the drought condition and corresponding yield of paddy (district-level) during Kharif 2018. Vegetation drought indices, namely Vegetation Condition Index (VCI), Temperature Condition Index (TCI), Vegetation Health Index (VHI), and vegetation indices (VI) anomaly, were derived from different VI (i.e., NDVI, EVI) to assess the paddy health condition during drought year (2018) and non-drought year (2017). Later, the correlation between the DES-based yield data and derived drought indices (for the year 2017) were made to develop the district-level paddy yield model for the drought year 2018. The key results of the study shown that VCI derived from EVI data was found to be more reasonable to depict the drought condition, wherein ~21% area was under severe drought condition, 43% area under moderate drought condition, and 36% area under no drought condition. In addition, the yield prediction model derived from VCI (EVI-based) was found to be promising for predicting the paddy yield for Kharif 2018 with fair R2 of 0.53.
Top1. Introduction
Agriculture and allied sectors have a significant contribution to the development of countries economy, wherein, it provides nutritious need to the billions of the world’s populations. During last few decades, the agriculture sector has been dramatically affected by the climate change events e.g. unseasonal rainfall, drought, hailstorms, strong wind surges, etc. (Patel et al., 2012; Dutta et al., 2015). These undesirable weather event poses loss in agriculture production; and subsequently, it leads to many socio-economic issues, such as, escalation in global commodity values, remortgage or sale of productive croplands to endorse the real estate industry, migration of peoples from rural area to municipal centre, increases the rashness and social offence, etc. (Vaani and Porchelvan., 2018).
Agricultural drought is one of the serious and undesirable events, which poses huge loss in agricultural productivity (especially during Kharif season) (Dalezios et al., 2014). Basically, drought is a phenomenon of a long period of dry conditions where a region observes a deficit in its water supply, whether surface or underground water. Droughts are recurring phenomena in India owing to moisture stress; and comes under the chronically drought-prone areas, and they receive less than ~750 mm rainfall, while, 35% of the region comes under the drought-prone area, and they receive ~750-1125 mm rainfall. Altogether, ~68% of the country falls under the drought-prone region and experiences severe drought conditions (PACS Programme 2001-2008). The agricultural drought badly affects the economy of many agrarian dominant countries (i.e. India, China, Bangladesh, Vietnam), where more than 65% of people are dependent on the agriculture and allied sector. In India, about 16% of the agricultural area comes under the drought-prone region, wherein, ~50 million people are annually affected by the drought events (DAC 2009). However, continuous decline in the share of agriculture and allied sectors in the Gross Value Added (GVA) is observed from 18.6% (during 2013-14) to 17.4% (during 2016-2017) (DAC annual Report 2017-18).
Giving importance to the agriculture sector, it is very crucial to have detailed and timely information on agricultural drought so that the impact of the drought can be mitigated by making the appropriate decision. Albeit, it is quite hard to acquire detailed, timely, and accurate information on drought by traditional survey-based methods. Typically, survey-based methods are cumbersome, required large man-power, costly, and large time-taking. In this line, remote sensing (RS) technology and Geographical Information System (GIS) have proven its effectiveness to monitor the agriculture system by its specific capabilities of spectral, temporal, radiometric, and spatial characteristics (Murthy et al., 2007; Ranjan et al., 2016a, 2016b). The RS-based methods provide a cost-effective, robust, reliable, easy-access, and time-saving aspect to monitor, mapping, and policy-making opportunities (Ranjan and Parida, 2019).
Basically, RS-based methods are used for crop acreage mapping and yield estimation, cropping pattern analysis, crop water management, agricultural drought assessment and monitoring, horticultural enlargement, precision agriculture farming, soil types mapping, watershed management and monitoring, assessment of climatic events on agriculture, and so on (Navalgund and Ray, 2000; Panigrahy and Ray, 2006; Navalgund et al., 2007; Parida and Ranjan, 2019a). In this regard, wide-range of study has been conducted over the globe using remote sensing technique to monitor the agriculture system (Ji et al., 2003; Jain et al., 2015; Sholihah et al., 2016). Nandeesha and Ramu (2015) conducted a study using 250m resolution MODIS TERRA images over central dry agro-climatic zone to assess the agricultural drought situation. Further, Sahu and Patel (2016) have utilized MODIS Normalized Difference Vegetation Index (NDVI) product from 2002-2015 in Bundelkhand region (India) for monitoring agricultural drought. Vaani and Porchelvan (2018) have also used the long-term NDVI of Global Inventory Modeling and Mapping Studies (GIMMS) for the period of 20 years (1984-2003) to derive Vegetation Condition Index (VCI), and thereby, to recognize the vegetation vigour and the fortnightly deviation of VCI during major crop growing period (June to September). Das et al., (2017) utilized the MODIS-based NDVI product for the years 2006 (Kharif season) and 2010 (both Kharif and Ravi season) over Puruliya district, West Bengal to assess the agriculture drought.