Assessment of the usefullness of radar inputs to the type of distributed numerical models planned for catchment management in Ireland

Eutrophication of lakes and rivers is a major concern in Ireland at the moment and phosphorous enrichment, from agricultural point and non-point sources, has been identified as a major factor. Research underway to assess the numerical hydrological modelling requirements for understanding and managing the problem at the relevent space and time scales typical of hydrological response in Ireland. Spatially distributed models which estimate soil erosion and water contamination, such as BASINS and SWAT, are included in these studies.

In Ireland, the problem is compounded by a high degree of spatial and topographical heterogeneity and relatively sparse raingauge network in agricultural areas. A major unanswered question is whether this limits the practical usefullness of typical distributed catchment models and, if so, whether other data sources, such as radar estimates of precipitation, can help.

This paper reports on a case study of the sensitivity of such hydrological models to the additional information about precipitation available from radar. It describes a case study on a small (110 km^2), hilly (sea-level to 400 m), agricultural catchment (Dargle) on the East Coast of Ireland. Radar precipitation information is provided from a C-band radar located at Dublin airport and there are 4 tipping bucket recording raingauges in the catchment and 8 water level recorders at various locations on the channel network to estimate discharge. The 8 water level recorders allow the catchment to be split into subcatchments representative of specific physical conditions. A severe (over 100 mm in approximately 12 hours )rainstorm occurred on the 5 and 6 November 2000. The effect of the additional radar information on the hydrological model predictions is assessed for this extreme event and compared with other, less severe events. Anticedent rainfall is an important determinant of the flood response.