Agricultural Drought Prediction Maps for Maize Fields in Eastern Austria, 2018-2023

Omid Ghorbanzadeh's dataset provides pixel-level drought maps for maize fields in the Marchfeld region of Austria, generated from Sentinel-2 satellite time series. The data includes over 20 satellite-derived indices integrated with ground data on field capacity, precipitation, and soil composition. Prediction models include Random Forest, Deep Neural Network, and 1D-CNN, with accuracy evaluated from 2018 to 2023.

Use Cases

• Compare machine learning model performance (RF, DNN, 1D-CNN) for drought prediction based on the described framework.
• Analyze drought signal differences between irrigated and non-irrigated agricultural fields based on the study's comparative analysis.
• Validate satellite-derived drought indices with ground data on field capacity and soil composition as described.
• Develop operational drought monitoring frameworks using high-resolution Sentinel-2 time-series data as outlined in the study.

Strengths

• Includes over 20 satellite-derived indices integrated with ground data.
• Model performance is evaluated over a multi-year period from 2018 to 2023.
• Covers both irrigated and non-irrigated fields for comparative analysis.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment.
• Data is focused on a specific case study in Eastern Austria, which may limit generalizability.

Provenance

Source

figshare, authored by Omid Ghorbanzadeh.

Collection Method

Satellite-derived indices from Sentinel-2 time series combined with ground data, processed through ML models.

Time Range

2018-2023

Freshness

Last updated 2026-04-13 07:10:48; freshness should be verified.

Geography

Marchfeld region, Eastern Austria.