Predicting monthly streamflows and their variability from limited historic records in the Tárcoles watershed, Costa Rica

Abstract

Lack of instrumentation in many developing nations makes it difficult to gauge available water resources, an essential part of national socioeconomic wellbeing. Costa Rica is a developing country that has considerable hydrologic data. This study establishes a simple model from limited data to predicted mean monthly flows and their interannual variability at unmonitored locations, within the densely populated Tárcoles watershed of Costa Rica. Relationships between basin area and percentiles of historic flows are derived as a matrix of monthly flow percentiles and averages that can be transformed into potential flows. The station used to validate the model has limited records and contains apparently abnormal flows, nonetheless results indicate considerable success of the model. It could further be refined and tested with more discharge stations. In turn, the modelling approach could be used to predict monthly flows and their variability within the Tárcoles basin as well as other countries where information may be lacking.