Abstract

GUEVARA OCHOA, Cristian et al. Filling series of daily precipitation for long periods of time in plain areas: Case study superior basin of stream del Azul. Geoacta [online]. 2017, vol.42, n.1, pp.38-62. ISSN 1852-7744.

The filling of missing daily precipitation data is a common problem in hydrological studies. The aim of this article is to make a comparison and evaluation of different methods to fill in missing daily rainfall data for long periods in plain areas. This study is carried out in the upper basin of the Azul stream for a period of nine years (2006-2014) and three stations that have 3287 complete data of daily precipitation and six stations with incomplete data are used. Seven methods were implemented for the filling of daily rainfall data: the linear regression method (MRL), the distance reasons method (MRD), the coefficients correlation method with neighboring stations (MRC) average reason method (MRP), the reasons distance method (MRD), the inverse distance weighted method (MIDW), following Markov chain method (MKV) and finally neural networks method (MRN). For the comparison and analysis of the different methodologies, different statistics and temporal graphs were applied, which measure the adjustment of the calculated data. Probabilistic and neural networks are the most suitable methods to fill data in plain areas. The methods applied in the study obtained a better adjustment in the autumn-winter season with lower rainfall, compared to the spring-summer period where lower adjustments were obtained because to that in these times there are convective storms with very high rainfall intensities.

Keywords : Filling data series; Daily Precipitation; Plain.

        · abstract in Spanish     · text in Spanish     · Spanish ( pdf )