Relación entre datos metereológicos y de sensores remotos entre los años 2000 y 2018 en Quito, Ecuador

Abstract

The surface temperature analysis is also known as (LST), it is one of the most relevant variables for established environmental studies that allow us to know the factors such as energy balance, evapotranspiration and soil moisture. For the present investigation, data from two sources were used, the first of satellite images of the Landsat 7 sensors, in the years from 2003 to 2018, and Landsat 8 between 2013 to 2018 respectively and in the second meteorological data, of the automatic stations of the Metropolitan Network of Atmospheric Monitoring of Quito. Once the satellite images were downloaded, they sometimes went to radiometric and atmospheric corrections in the ENVI 5.0 program to minimize errors that these losses. Subsequently, the corrected images are loaded into a Geographic Information System, specifically ArcGIS 10.6 from which a series of environmental indicators are obtained, with which a multivariable matrix is calculated. In the modeling phase the data were treated for years, decades and combination of variables generating 3 linear models for each sensor, the most relevant models that passed to the validation stage obtained a correlation coefficient or also called R2 of 0.8329, R2 Adjusted 0.7661 for Landsat 7, while Landsat 8 obtained an R2 of 0.8117 and an Adjusted R2 of 0.7363 for the year 2014 in the two sensors, thus concluding that the Maximum, Average and Minimum Temperature they are the variables most related to the LST.