Cumulative Impact Analysis of Rain Gauge Networks on Rainfall Forecasting Accuracy: A Case Study

Accurate rainfall estimation is vital for real-time forecasting of water resources used for water demand management. This needs an optimum density of rain gauges. While numerous geostatistical methods exist for optimizing rain gauge networks, many may suffer from limitations. This study aims to develop a novel geostatistical method to redesign rain gauge networks that was demonstrated in a real-world case study of Khorasan Razavi province, specifically in the Qarahqoom basin, Iran, aiming to minimize errors. The methodology involves analyzing the number and locations of rain gauges and assessing each rain stations contribution to the regions overall rainfall estimation accuracy. Initially, station homogeneity in the study area is verified using the linear moment method. Subsequently, a suitable semi-variogram is selected to calculate the acceptance probability for different areas within the province. This approach determines acceptance accuracy (AP) values at various probability levels. Considering the basins characteristics, including its homogeneity, the acceptance probability method was implemented at an 80% probability level. The findings reveal that current networks of 66 rain gauges achieves a 61% acceptance accuracy. Of these, only 42 rain gauges significantly influence the estimated basin rainfall (i.e. forming the base network) while the remaining 24 rain gauges have a minor impact (i.e. non-base network). It is proposed that adding 24 strategically placed stations could evaluate the rainfall estimation accuracy in the Qaraqoom basin to 95%. Keywords: Rain gauge network, variogram, acceptance probability, acceptance accuracy, Qarahqoom basin