TY - JOUR AU - Rangel Sotter, Adriana Patricia AU - Sánchez Ipia, Alber Hamersson AU - Siabato Vaca, Willington Libardo AU - Cely Pulido, John William PY - 2001/11/30 Y2 - 2024/03/29 TI - Geoestadística aplicada a estudios de contaminación ambiental JF - Ingeniería JA - Ing. VL - 7 IS - 2 SE - Science, research, academia and development DO - 10.14483/23448393.2815 UR - https://revistas.udistrital.edu.co/index.php/reving/article/view/2815 SP - 31-38 AB - Some people say that Geostatistics is "the art of modeling spatial data". A more specific definition is: "Geostatistics is a statistical methodology used to estimate, forecast, and simulate correlated spatial data, which uses in its analysis exploratory and interpolative methods2" Geostatistics is a useful tool for improving estimations of a variable for non-measured locations if it is compared with other estimation techniques, for example, IDW (Inverse Distance Weighted interpolation). Using Geostatistics for modeling air pollutants behavior in Bogotá has two immediate advantages: It is not necessary to collect as much information as it is with the Gaussian Plume Model. The exploratory analysis used the Bogotá's air quality network data (14 stations taking 24 air pollutant measures a day) for a first approach. It is not mandatory to collect information about each air pollutant source. For this reason, it is possible to draw some conclusions faster than with other approaches. A preliminary analysis conclusion was obtained through the use of cross validation to compare IDW against Geostatistics (Ordinary Kriging) RMSE (Root Mean Square Error). This conclusion applies for Bogotá case and says that it is better to use IDW instead of Geostatistics for estimating air pollution based on the available sample points (air quality stations). The research team thinks that this conclusion is the direct result of the lack of sample points. Based on this Geostatistical analysis, the research team realized that DAMA must to extend and crowd the station network for better future estimations. ER -