Remote Sensing and GeoAI Methods for Wildfire Impact Assessment in Mediterranean and Similar Ecosystems

Description

This dataset contains structured information extracted from peer-reviewed journal articles included in a systematic review on remote sensing and GeoAI applications for post-fire impact assessment in Mediterranean and Mediterranean-type ecosystems. The dataset compiles study-level attributes, including publication information, study area and ecosystem context, sensor types and platforms, spatial and temporal resolution, target variables, application purposes, methodological categories, algorithm families, data levels, input feature types, and validation strategies. The data were collected through a systematic literature search and manual data extraction following predefined inclusion and exclusion criteria and the PRISMA 2020 protocol. The dataset is intended to support reproducibility, comparative analyses, and future methodological and bibliometric studies on wildfire impact assessment and GeoAI-based environmental monitoring.

Steps to reproduce

The dataset was created through a systematic literature review conducted in accordance with the PRISMA 2020 guidelines. Peer-reviewed journal articles were identified through structured searches in the Web of Science and Scopus databases using predefined keyword combinations related to wildfire impacts, remote sensing, and GeoAI applications. Following the initial retrieval, duplicate records were removed and titles and abstracts were screened to assess relevance. Full-text screening was subsequently performed using predefined inclusion and exclusion criteria focusing on post-fire impact assessment in Mediterranean and Mediterranean-type ecosystems. Studies addressing only fire risk, fire occurrence, or early warning systems without post-fire impact analysis were excluded. For each included study, relevant information was manually extracted into a structured data extraction form. The extracted variables include bibliographic information, study area and ecosystem context, sensor and platform types, spatial and temporal resolution, target variables and application purposes, methodological categories, algorithm families, data levels, input feature types, and validation strategies. Data extraction and curation were carried out using Microsoft Excel, and the final datasets were exported in comma-separated values (CSV) format. Aggregated supplementary tables were derived by grouping and summarising the study-level variables. No automated text-mining or machine-reading tools were used; all records were reviewed and annotated manually to ensure consistency and accuracy.

Institutions

• Işık University

  Istanbul, Istanbul

Categories

Licence