Estimation of missing data and maps of the behaviour of total precipitation in northern colombia

Abstract

The climatic conditions of a territory are a key element in the development of multiple types of projects, hence it is necessary to have relevant information that helps determine and simulate the future climate conditions that may influence such projects. At this point, the important thing is to have reliable and complete weather information records. Unfortunately, in Colombia, being able to find this type of information is not so simple, and in general the climatological data are incomplete for reasons such as failures or non-calibration in the measurement instruments, errors in the recording of measurements, meteorological extremes and difficulty in the access of the measurement areas. Hence, the determination of the missing data in the records of the climatological stations becomes a necessary question to complete the incomplete historical series. A common case is the missing data in rainfall stations, which are essential in hydrological simulation studies in order to obtain optimal hydraulic solutions. Therefore, it is almost mandatory to properly fill in missing data before performing any analysis on such data. This work presents the determination of missing data of total precipitation of 40 meteorological measurement stations located in the north of Colombia. The series analysed cover a period of 40 years of observations from 1980 to 2019, obtained by the Instituto de Hidrología, Meteorología y Estudios Ambientales (IDEAM). The estimation of missing data was carried out using Multiple Linear Regression (MLR), which allows estimating missing data based on nearby records that correlate with said data. The Beale - Little algorithm is applied which is based on the RLM method, using all the available information and leads to a simultaneous estimation of the missing values. The estimated precipitation dataset is useful for future hydrological analysis of studies that require meteorological forecasts in regions with losses of hydrological information, such as the meteorological stations located in the north of Colombia.