Spatial Database of Precipitation of Rio Grande do Sul

Abstract

Currently, when getting time series of precipitation, it is necessary to perform gap filling, interpolate and estimate precipitation for the study área, mainly in areas where there is a low density of rainfall stations. This study aimed to generate a spatial database with historical time series of precipitation for the state of Rio Grande do Sul, that allows the consultation of precipitation indexes by watershed, by municipality and by geographic coordinates. The methodology was structured in five stages: acquisition, organization and gap filling in time series of precipitation; interpolation, by the method Inverse Distance weighting (IDW), rain data for a regular grid with spatial resolution of 20 km; calculate precipitation rates (precipitation return, average monthly and annual rain, Rainfall Anomaly Index (RAI), number of rainy days); coding and dividing river basins from the Digital Elevation Model (DEM); generate database: organize matrices and tables to consult. The cross validation of the interpolation presented Mean Absolute Error (MAE) that varied between 1,02 and 3,20 mm, while the Mean Square Error (MSE) varied between 6,4 and 8,4mm. The availability of this database on the internet, with an output file compatible with most GIS software, represents an important gain in research that needs to use long time series.

Keywords
historical time series; gap filling; spatial interpolation; Pfafstetter code