Skip to content

yuyi13/dSMADI

Repository files navigation

A downscaled soil moisture agricultural drought index (dSMADI) through spatiotemporal fusion

License: MIT R 4.3.1 Status Preprint

Contents

Overview

This repository is structured to support reproducible research for "Assessing how spatiotemporal fusion enables an agricultural drought index to represent ecosystem evaporative stress dynamics at continent-to-field scales" (Yu et al., 2026). The study investigates how a modified Soil Moisture Agricultural Drought Index (SMADI) can be translated into a downscaled product (dSMADI) that links continental drought monitoring with field-scale ecosystem evaporative stress dynamics across Australia.

At this stage, the repository is a documentation-first scaffold. The README and directory layout define the intended workflow, while scripts, figures, and ancillary assets are represented with placeholders until the implementation is added.

Repository structure

The current repository layout is:

dSMADI/
├── 0_ancillary/                  Ancillary inputs, metadata, and external-data notes
├── 1_smadi_design/               SMADI concept, formulation, and design placeholders
├── 2_experimental_scripts/       Four-step dSMADI workflow placeholders
├── 3_figure_scripts/             Figure reproduction placeholders
├── figures/                      Exported figure placeholders for README-linked assets
├── preprint/                     ESSOAr manuscript PDF
├── LICENSE                       MIT license
└── README.md                     Project overview, workflow summary, and citation details

Background

Remote-sensing drought indices are widely used for agricultural drought monitoring, but coarse spatial resolution can obscure within-field stress heterogeneity. The dSMADI study addresses this by combining climatology-removed soil moisture, land surface temperature, and lagged vegetation information, then using spatiotemporal fusion to produce a 100 m weekly drought index that can be compared with ecosystem evaporative stress observations.

The workflow spans continental-to-field scales across Australia during 2016-2023 and focuses on four linked objectives: baseline SMADI generation, downscaling through fusion, evaluation against independent stress observations, and variance decomposition of component contributions.

SMADI design

The directory 1_smadi_design/ replaces the single demo-style entry point used in OzNet_AOA. It is intended to hold:

  • conceptual notes on the original SMADI formulation;
  • placeholders for how soil moisture, land surface temperature, and NDVI are combined;
  • future worked examples or diagrams explaining the progression from SMADI to dSMADI.

Current status:

[placeholder for future SMADI design notes]

Experimental scripts

The directory 2_experimental_scripts/ will follow the same numbered workflow idea used in OzNet_AOA, but mapped to the four objectives of the dSMADI manuscript:

  1. Generate baseline SMADI across Australia and compare it with reference drought metrics such as SPEI.
  2. Apply spatiotemporal fusion to produce weekly 100 m dSMADI.
  3. Evaluate dSMADI against ECOSTRESS ESI and OzFlux flux tower observations.
  4. Quantify the contributions of the SM, LST, and NDVI components through variance decomposition.

These workflow components are placeholders for now and are not yet implemented as runnable scripts.

Figure scripts

The directory 3_figure_scripts/ is reserved for future figure-by-figure reproduction scripts and shared helpers. The corresponding exported assets will live under figures/ once they are available.

Current status:

[placeholder for future figure scripts and exported figures]

Preprint and data

The manuscript currently included in preprint/ is:

  • Yu-2026-ESSOAr-Preprint-15001424v1.pdf

Preprint DOI:

Data links and permanent archival records will be added here when they are ready.

Current status:

[placeholder for future data and permalink details]

How to cite

@article{Yu2026_dSMADI_ESSOAr,
  author = {Yi Yu and Tim R. McVicar and Thomas G. Van Niel and Luigi J. Renzullo and Siyuan Tian and Brendan P. Malone and Jian Peng and Patrick Filippi and Thomas F. A. Bishop},
  title = {Assessing how spatiotemporal fusion enables an agricultural drought index to represent ecosystem evaporative stress dynamics at continent-to-field scales},
  journal = {ESS Open Archive},
  year = {2026},
  doi = {10.22541/essoar.15001424/v1}
}

About

A downscaled soil moisture agricultural drought index through spatiotemporal fusion

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors