Trend analysis of an historical series of flow and precipitation using a non-parametric statistical test

Abstract

Detecting trends in flow and precipitation series has great importance for the management of water resources and can be accomplished using statistic non-parametric tests. This work evaluated the significance of trends in flow and temporal series of precipitation as well as monthly and annual precipitation trends using Pettitt’s test. The study used historical series of flow and precipitation corresponding to the period of 1939 to 2005, recorded at stations 02041003 and 57450000 of the hydro-meteorological network of Water National Agency (ANA), located in the municipality of Alegre-ES. Using the Pettitt test, it was possible to identify a change in the trend of October average flow since the year 1963. The average for this month increased 34.2% since 1963. However, the rainfall series had a positive trend in August and September since the years 1967 and 1964, respectively. The increase in the October average flow may be a consequence of an increase in rainfalls in August and September; this requires further studies which may consider the influence of climate and anthropic factors in this hydrologic cycle component. Based on the results of the study, it was concluded that the Pettitt test performed satisfactorily in detecting trends in the historic series of flow and precipitation.