# READ ME

Dataset title: Surface Water Transitions 1984–2022: A Global Dataset at Annual Resolution
Dataset DOI: [https://doi.org/10.6084/m9.figshare.28138643](https://doi.org/10.6084/m9.figshare.28138643)
ReadMe Author: Gustavo Willy Nagel, University of Southampton
ORCID ID: [https://orcid.org/0000-0002-8060-0772](https://orcid.org/0000-0002-8060-0772)

This dataset supports the thesis entitled: Global Detection and Analysis of Surface Water and River Meander Dynamics using Cloud-Based Remote Sensing

AWARDED BY: University of Southampton
DATE OF AWARD: December 2025

Date of data collection: 1984–2022 (Landsat time series)

Information about geographic location of data collection:
Global (worldwide coverage of rivers, lakes, reservoirs, wetlands, and other surface water bodies).

Licence:
Creative Commons Attribution 4.0 International (CC BY 4.0)

Related projects/Funders:
INSPIRE Doctoral Training Programme (DTP) funded by the Natural Environmental Research Council (grant number NE/S007210/1)

---

DATA & FILE OVERVIEW

---

This dataset contains:

1. Water Transition Data (Raster)
   Global raster images derived from Landsat time series (1984–2022) indicating surface water transitions.

Band 1 (b1): Water advance
Band 2 (b2): Water recession
Each pixel records whether surface water expanded or receded and the year in which the transition occurred.

2. Statistical Summary (GeoJSON)
   A spatial vector file containing aggregated statistics of surface water transitions for specific regions or water bodies, including the magnitude and direction (advance or recession) of changes.

3. Processing Code (TXT)
   Google Earth Engine (GEE) JavaScript code used to process Landsat imagery and generate the water transition products.

Relationship between files:
The raster water transition files provide pixel-level information on surface water dynamics. The GeoJSON file summarizes these raster-derived transitions spatially for defined regions or water bodies. The GEE JavaScript file documents and enables reproduction of the full processing workflow.

Additional related data not included:
Raw Landsat imagery is not included and must be accessed directly via the Google Earth Engine data catalogue.

Source data:
Landsat satellite imagery (NASA/USGS), accessed and processed via Google Earth Engine.

Versioning:
Single version released. No previous versions.

---

METHODOLOGICAL INFORMATION

---

Description of methods used for data collection/generation:
The dataset was generated using annual Landsat surface reflectance time series from 1984 to 2022. Water detection and temporal analysis were performed to identify persistent and changing surface water pixels. Transitions were classified as water advance or water recession, and the year of transition was recorded.

Methods for processing the data:
Landsat imagery was processed within Google Earth Engine using a JavaScript workflow. Annual composites were generated, water masks were derived for each year, and interannual comparisons were used to detect transitions. Outputs were exported as global raster datasets and summarized spatially into GeoJSON statistics.

Software and instruments:
Google Earth Engine (cloud-based geospatial processing platform)
Landsat missions (Landsat 5, 7, 8)

Standards and calibration:
Standard USGS Landsat radiometric and geometric calibration and surface reflectance products were used as provided within Google Earth Engine.

Environmental/experimental conditions:
Not applicable (satellite remote sensing analysis).

Quality assurance procedures:
Quality control relied on established Landsat preprocessing, cloud masking, and temporal consistency checks across the annual time series.

People involved:
Dataset creation, processing, and submission by the author.

---

DATA-SPECIFIC INFORMATION

---

Water Transition Raster

Number of variables: 2 bands
Number of cases/rows: Pixel-based raster (global coverage)

Variable list:
b1: Water advance (annual transition year)
b2: Water recession (annual transition year)

Units: Year (integer)

Missing data codes:
NoData pixels indicate no detected transition or non-observable areas.

Statistical Summary (GeoJSON)

Variables:
Aggregated statistics describing magnitude and direction of surface water transitions by region or water body.

Specialized formats:
GeoTIFF (raster), GeoJSON (vector), TXT (code).

Date file created:
2025

---

NOTES

This dataset is intended for researchers, policymakers, water resource managers, and conservation practitioners. It supports analyses of river dynamics, wetland change, floodplain processes, and long-term impacts of climate change and human activities on global surface water distribution.