Statistical spatial and temporal analysis of historical rainfall and evapotranspiration behavior for Ecuador

Abstract

Statistical analysis of temporal and spatial data: example of rainfall, temperature and evapotranspiration in Ecuador Many studies in Ecuador had taken the endeavor of climate characterization, mainly referred to rainfall and temperatures, and some others to estimating evapotranspiration based on data availability. Such studies have only had local application and had focused on the rainfall gauging stations; fewer stations that have temperature records and even lesser that have information of the various parameters required for the widely used Penman - Monteith equation for estimating evapotranspiration. The main challenge in modeling climate in Ecuador comes from its location in the Intertropical convergence zone, the marine currents in the Pacific, the Amazon basin and the Andes. Therefore, characterization of climate is a first step in order to understand its spatial and temporal variability, and from here to undertake the task of generalizing meteorological elements to the whole country. There needs to be a compromise between data availability that sets boundaries to what can be done and the research objectives of spatial climate patterns. Regarding information, monthly data is available referred to rainfall, for a lesser number of stations temperature is also available; and other elements like wind speed, relative humidity, dew point, pressure, and solar radiation, are only available for a selected set of stations, making it difficult to count on them. GIS spatial data features are also available, like the SRTM DEM 90m, from which morphological characteristics can be derived. First step towards analysis is data validation. Many unforeseen circumstances affect meteorological elements measure and record until final presentation of raw data, and one of the most widely used technique to detect systematic errors is the double mass curve analysis, which will be applied mainly to rainfall information, time series plots may also help identifying errors in other elements like temperatures. Spatial distribution of rainfall will be addressed by means of distribution and seasonal rainfall patterns that will enable categorizing climatic regions. Correlation analysis will be used to estimate missing values and also to estimate ungauged parameters. The main aim is to contribute to a better understanding of climate in Ecuador; by means of models including Andes mountain range and Amazon and coastal regions, for the estimation of various elements like rainfall, temperature and evapotranspiration. Spatial regions with different behavior in rainfall, temperatures and evapotranspiration will be identified. Last but not least, since Ecuador is heavily affected by the extreme anomalies of ENSO, whose presence causes long term heavy rainfalls, and floods, climate characterization should consider its influence.