Análise de interpoladores espaciais para o mapeamento de chuvas mensais na Bacia Hidrográfica do Rio Ijuí

Abstract

In recent decades, research related to water scarcity has increased significantly, mainly due to climate change, which worries a large part of the scientific community. In this context, the control and study of the variables that are part of the hydrological cycle becomes increasingly relevant. However, there is a lack of reliable hydrological data, since there is a large number of flawed data and few on-site measurement stations. Therefore, it is important that ways to circumvent the scarcity of hydrological data are widely explored. In this sense, the use of spatial interpolation makes it possible to circumvent the problem of spatial continuity of hydrological data. For a more efficient management of information, the evaluation of different interpolators allows the best spatial interpolator to be chosen, for spatialization of a given phenomenon, in this work, precipitation. Spatial interpolations can be performed in geographic information systems software such as ArcGIS and QGIS. Therefore, in addition to choosing the best technique for spatial interpolation of rainfall, determining which of the two most used GIS software in the world has the best performance is also relevant. Therefore, this work consisted of spatializing accumulated monthly rainfall over the perimeter of the Ijuí river basin, located in the Pampa and Atlantic Forest biomes, in the state of Rio Grande do Sul. The interpolations were carried out for the months of January and August of the years 1990, 2000, 2010 and 2020. Using eight spatial interpolators, four processes were carried out in QGIS, these being the Weighted Angular Distance (ADW), inverse of the weighted distance (IDW), Irregular Triangle Network (TIN) and Ordinary Kriging (OK) with exponential variogram; the other four processes were performed in ArcGIS, these being the Natural Neighbor (Nat), Cubic Spline (Spline), Top to Raster (Top) and Trend (TREND) methods. In order to validate the efficiency of the different interpolators, the spatialization results were compared with information from the Climate Hazards group Infrared Precipitation with Stations (CHIRPS) database, which has high spatial resolution, in the case of global precipitation databases. This comparison was performed using the Willmott Agreement Index (d), Root Mean Square Error (REMQ), Mean Absolute Error (AME) and Mean Absolute Percentage Error (EPAM). The work concluded that the best spatial interpolator was IDW, which is present in both software. Finally, data consisting of spatial maps of the location of monthly accumulated rainfall data, in a specific format, with a spatial resolution of 0.25km².