{"id":761,"date":"2024-11-22T06:00:02","date_gmt":"2024-11-22T11:00:02","guid":{"rendered":"https:\/\/staging-colombia.mapbiomas.org\/?page_id=761"},"modified":"2025-11-25T08:21:46","modified_gmt":"2025-11-25T13:21:46","slug":"metodologia-mapbiomas-agua","status":"publish","type":"page","link":"https:\/\/colombia.mapbiomas.org\/en\/metodologia-mapbiomas-agua\/","title":{"rendered":"Method MapBiomas Water"},"content":{"rendered":"

WATER SURFACE MAPPING: METHOD SUMMARY<\/strong><\/p>\n\n\n\n

Here we present a summary of the MapBiomas Water Method. Access the ATBD (Algorithm Theoretical Basis Document ) in this LINK<\/a> for more methodological details. .<\/strong><\/p>\n\n\n\n

Presentation<\/strong><\/p>\n\n\n\n

The main objective of MapBiomas Water is to map surface water dynamics throughout the territory of Amazonian countries (Pan-Amazon), on a monthly and annual basis from 1998 to 2024. The datasets are publicly available on a web platform to improve the management and use of water resources across the Pan-Amazon region.<\/p>\n\n\n\n

The project used Landsat images from Google Earth Engine (1998\u20132024) to map water surfaces in Colombia. After dividing the country into 133 zones, spectral mixture analysis was applied to break down the pixels into pure components. These results were fed into a Random Forest classifier trained with MapBiomas samples. Post-classification included: masking with spectral indices (MNDWI, NDVI, AWEI, etc.), filling monthly gaps, filtering by temporal frequency, and geometric cleaning with reference layers. The results distinguish between permanent water (\u22656 months\/year) and seasonal water (1-5 months\/year).<\/p>\n\n\n\n

El mapeo comprendi\u00f3 el periodo de 1998 a 2024 en escala mensual, dado que a partir de 1998 la calidad de los mosaicos satelitales y por tanto de la detecci\u00f3n de agua superficial presenta niveles \u00f3ptimos para an\u00e1lisis rigurosos. Durante este per\u00edodo se procesaron y analizaron m\u00e1s de 396.000 escenas Landsat en la plataforma Google Earth Engine, estableciendo una base de datos robusta y confiable para el monitoreo de recursos h\u00eddricos en la regi\u00f3n amaz\u00f3nica.<\/p>\n\n\n\n

Organization and database\u00a0<\/strong><\/p>\n\n\n\n

The overall coordination of MapBiomas Water is led by Imazon and RAISG, while technical and operational coordination is directed by Geokarten. The reconstruction of the monthly historical series of surface water was carried out by specialists from all biomes of the Amazonian countries, under the leadership of the following institutions: Fundaci\u00f3n Amigos de la Naturaleza -FAN- (Bolivia), Fundaci\u00f3n Gaia Amazonas -FGA- (Colombia), EcoCiencia (Ecuador), Instituto del Bien Com\u00fan -IBC- (Per\u00fa), Provita y Wataniba (Venezuela), Alliance of Bioversity International y CIAT (Guianas y Suriname). The surface water mapping algorithm was developed by Imazon and adapted by MapBiomas Water in this initial phase of work.<\/p>\n\n\n\n

The development of the MapBiomas Water control panel (dashboard) was conducted by Geodatin and includes significant contributions from the MapBiomas Water working group and platform users in the design thinking process.<\/p>\n\n\n\n

Three types of products were produced by MapBiomas Water:<\/p>\n\n\n\n

    \n
  1. Monthly and annual surface water maps;<\/li>\n\n\n\n
  2. Surface water transition maps between \u201cWater\u201d and \u201cNon-water\u201d classes. This product was processed using the annual surface water database;<\/li>\n\n\n\n
  3. Trend maps (increase and decrease) in surface water. This product was calculated from monthly surface water data in 5 km x 5 km grids. <\/li>\n<\/ol>\n\n\n\n

    The dashboard (link<\/a>) consists of maps, statistics, and visualization, analysis, and data access tools. It is possible to view the data on an annual and monthly scale, as well as obtain it in different territorial units. Finally, the dashboard also provides a link to access the MapBiomas Water data API. <\/p>\n\n\n\n

    Method<\/strong><\/h4>\n\n\n\n

    1. Pre-processing<\/strong><\/p>\n\n\n\n

    The process began with the acquisition of Landsat satellite images from missions 5, 7, 8, and 9, corresponding to the TM, ETM+, and OLI sensors, available on Google Earth Engine. Specifically, Collection 3 Level 1 orthorectified images with a spatial resolution of 30 metres were used. The analysis period covered 1998 to 2024, applying a filter to select only scenes with less than 70% cloud cover over Colombian territory.<\/p>\n\n\n\n

    2. Regionalisation<\/strong><\/p>\n\n\n\n

    Given Colombia's size and geographical diversity, the country was divided into 133 working areas measuring 100x100 km\u00b2, allowing for more efficient processing tailored to regional characteristics.<\/p>\n\n\n\n

    \"\"<\/figure><\/div>\n\n\n\n

    <\/p>\n\n\n\n

    3. Spectral Mixing Analysis (SMA)<\/strong><\/p>\n\n\n\n

    This fundamental stage allowed the subpixel composition of the images to be characterised by identifying six main components: Green Vegetation (GV), Non-Photosynthetic Vegetation (NPV), Soil, Clouds, Shade and Snow. SMA was crucial for detecting surface water mixed with other components, overcoming the limitations of whole-pixel classifiers.<\/p>\n\n\n\n

    4. Classification with Random Forest<\/strong><\/p>\n\n\n\n

    A balanced sample set was generated using the MapBiomas Agua Colombia collection as a reference, stratified by year and month, which included representative examples of both classes: \"water\" and \"non-water\".<\/p>\n\n\n\n

    \"\"<\/figure><\/div>\n\n\n\n

    <\/p>\n\n\n\n

    Subsequently, the Random Forest algorithm was trained using the stratified samples, considering the regional proportion of each class. This approach ensured that the model was adapted to the specific characteristics of each area and time period.<\/p>\n\n\n\n

    \"\"<\/figure><\/div>\n\n\n\n

    <\/p>\n\n\n\n

    5. Post-Classification<\/strong><\/p>\n\n\n\n

    Thresholds were applied to multiple indices to refine the classification:<\/p>\n\n\n\n