RECONCILING PRECIPITATION WITH RUNOFF:
THE ROLE OF UNDERSTATED MEASUREMENT BIASES IN THE MODELLING OF HYDROLOGICAL PROCESSES
Inaccurate precipitation measurements have been recently recognised as the “wilfully neglected Achille’s heel” of hydro-meteorological sciences. Difficulties in achieving accurate measurements arise from various instrumental and environmental sources of systematic biases, resulting in a significant underestimation of the precipitation depth and intensity. The understated extent of the associated biases is largely unknown and varies with various environmental factors, due to the complexity of the controlling processes. Although attempts were made to standardise measurement procedures, this has never been successfully achieved.
Without any correction for or, in many cases, any awareness of such measurement errors, there is a grave risk of a breakdown in the understanding of hydro-meteorological processes in a scientific era dominated by modelling, which generally undervalues the principals of precise and accurate measurements. Implications describe an inconvenient truth in hydrological sciences, which transcends a variety of applications of precipitation data in hydrological models, from real-time flood forecasting to water resources management and urban hydrology. The calibration of satellite- and radar-based areal estimates of precipitation and the statistics derived from historic data series are also systematically affected.
The extent and implications of inherent instrumental biases and wind-induced undercatch of precipitation measurements in the modelling of hydrological processes at the basin scale is the main focus of this research project. The aim is to quantify the impact of incorrect measurements used as the forcing variable of physically based hydrological models on their typical output variables, including the flood peak and volume, time to peak, baseflow separation and the regression curves. The impact on the simulation of hydrological processes at the basin scale is investigated, such as evapotranspiration, infiltration, interception, etc. The methods used to achieve the project objectives include theoretical analysis, numerical simulation (CFD, distributed hydrological models, interpolation and data integration, statistical analysis) and full-scale experiments performed in the laboratory (wind tunnel) and in real-world experimental basins.
The main expected result is to provide scientific evidence of precipitation measurement biases and their impact on hydrological models, by showing the improvement obtained when corrections for instrumental and environmental errors are implemented. To achieve this, other intermediate results will be obtained, e.g. the development of suitable correction curves for the wind-induced undercatch, the improvement of areal rainfall estimates based on the integration of rain gauge, radar and satellite sources, and the development of dedicated statistical tools to improve the assessment of homogeneity in precipitation time series, climatic trends and extreme value statistics.