Methodology to Create Geospatial MODIS Dataset

Abstract

Training and testing of algorithms used in computing for application in several studies, require datasets previously validated and labeled. In the case of satellite remote sensing, there are several platforms with large volumes of open source data. Aqua and Terra satellite platforms have available the sensor MODIS (Moderate-Resolution Imaging Spectroradiometer) which has available open access data for earth observation. Despite the facilities offered by the MODIS data platform, extracting data from a particular region for the construction of useful dataset requires an arduous work that includes manual, semi-automatic and automatic stages. The present study proposes a methodology for the construction of a geospatial dataset using MODIS sensor data. This methodology has been successfully implemented in the construction of dataset for the analysis of physical and biological variables in the Caribbean Sea, highlighting its application in the monitoring of Sargasso along the coastline of the state of Quintana Roo. Its application can be extended to any of the data and products offered by the MODIS sensor.