Skip to main content
Drought Intelligence Explorer: a 45-year district-level dataset of ten drought indicators for India and an open analytical platform

Drought Intelligence Explorer: a 45-year district-level dataset of ten drought indicators for India and an open analytical platform

This is a Preprint and has not been peer reviewed. This is version 1 of this Preprint.

Add a Comment

You must log in to post a comment.


Comments

There are no comments or no comments have been made public for this article.

Downloads

Download Preprint

Authors

Ankit Deshmukh 

Abstract

Publicly available drought records for India rarely combine sub-national detail, multi-decadal coverage, several indicator families and a documented processing chain in one citable product. Here we present a monthly, district-level dataset of ten drought indicators covering all of India from January 1981 to December 2025, together with the open, browser-based platform that serves it. Four of the indicators are standardized z-score indices: the Standardized Precipitation Index (SPI) and the Standardized Precipitation Evapotranspiration Index (SPEI), each at 1-, 3-, 6- and 12-month accumulation windows. The other two are percentage-based indicators: a root-zone Soil Moisture Anomaly (SMA) and the Vegetation Condition Index (VCI). The dataset covers 724 administrative districts in 36 States and Union Territories and holds 3{,}715{,}048 valid district-month values. The indicators come from CHIRPS v2.0 precipitation, ERA5-Land potential evaporation and root-zone soil moisture, and MODIS MOD13A3 NDVI, extracted and aggregated by a scripted Google Earth Engine pipeline. SPI, SPEI and SMA use a fixed 1981--2020 reference period (40 years); VCI uses 2000 to the release date. Missing observations are never imputed. As a scientific consistency check, the released record reproduces documented drought episodes with the expected geography and timing (for example, every Rajasthan district is flagged in July to September 2002, and the vegetation index collapses two to three months after the 2015 Marathwada rainfall deficit). Software quality is guarded by per-fit calibration gates, a checksummed release manifest and an 82-assertion offline self-test. Every value is reproducible from public inputs and documented commands, citable through a versioned DOI (10.5281/zenodo.24113364), and analysable and exportable through the platform without an account or backend. The product is a transparent research instrument, not an operational drought declaration.

DOI

https://doi.org/10.31223/X5F52X

Subjects

Agriculture, Civil and Environmental Engineering, Engineering, Environmental Sciences, Environmental Studies, Geography

Keywords

drought; SPI; SPEI; Vegetation Condition Index (VCI); soil moisture anomaly; India; Google Earth Engine; reproducible research; district-level dataset; geospatial web platform

Dates

Published: 2026-11-11 22:01

Last Updated: 2026-11-11 22:01

License

CC BY Attribution 4.0 International

Additional Metadata

Conflict of interest statement:
None

Data Availability:
The dataset and software supporting this study are openly available. Archived release (versioned DOI): Zenodo, https://doi.org/10.5281/zenodo.24113364 (version 1.0.0, released 2026-11-10). Source repository: https://github.com/anixn/drought-intelligence-explorer (tag v1.0.0). Live platform: https://drought.ankitdeshmukh.com The archive contains the ten indicator matrices (SPI-1/3/6/12, SPEI-1/3/6/12, SMA and VCI) for 724 districts and 36 States/UTs from 1981-01 to 2025-12 (VCI from 2000-02), the district and State/UT boundaries, the region registries, the metadata.json data contract, two precomputed cross-check files, and the full processing pipeline and application source code. The software is released under the MIT licence and the archived record under CC BY 4.0. The input products keep their own licences: CHIRPS (public domain), ERA5-Land (Copernicus licence, attribution required) and MOD13A3 (NASA open data). District boundaries come from the community-maintained india-maps-data repository (https://github.com/udit-001/india-maps-data); that source states no explicit redistribution licence, so the boundaries are redistributed with attribution and its disclaimer, and the boundary layer can be replaced without affecting the indicator values. All values are reproducible from public inputs using the documented commands in the repository.

Metrics

Views: 16

Downloads: 1