The surface runoff process in fluvial/pluvial flood modelling is often simulated employing a two-dimensional (2-D) diffusive wave approximation described by grid based digital elevation models (DEMs). However, this approach may cause potential problems when using the 2-D surface flow model which exchanges flows through adjacent cells, with conventional sink removal algorithms which also allow for flow exchange along diagonal directions, due to the existence of artificial depression in DEMs. In this paper, we propose an effective method for filling artificial depressions in DEM so that the problem can be addressed. We firstly analyse two types of depressions in DEMs and demonstrate the issues caused by the current depression filling algorithms using the surface flow simulations from the MIKE SHE model built for a medium-sized basin in Southeast England. The proposed depression-filling algorithm for 2-D overland flow modelling is applied and evaluated by comparing the simulated flows at the outlet of the catchment represented by DEMs at various resolutions (50 m, 100 m and 200 m). The results suggest that the existence of depressions in DEMs can substantially influence the overland flow estimation and the new depression filling algorithm is shown to be effective in tackling this issue based upon the comparison of simulations for sink-dominated and sink-free DEMs, especially in the areas with relatively flat topography.

Multi‐source remote sensing precipitation products are being used more and more frequently for rainfall analysis and hydrological applications around the world. The aim of the present study was to assess the variances and uncertainty of long‐term precipitation and corresponding hydrological simulations from 1984 to 2016 using rain gauge measurements, two satellite rainfall products retrieved from the Climate Prediction Center (CPC) and the Precipitation Estimation from Remotely Sensed Information using Artificial Neural Networks – Climate Data Record (PERSIANN‐CDR) respectively, along with rainfall reanalysis data from the ERA‐Interim over the lower Yangtze River basin, China. The results indicated overestimation from ERA‐Interim and PERSIANN‐CDR on seasonal rainfall. CPC rainfall estimation had marginal bias with rain gauge measurements among all the remote sensing rainfall products on a daily scale, while the ERA‐Interim tended to overestimate when the rainfall rate was less than 50 mm. On a monthly scale, CPC rainfall estimations were very close to the rain gauge measurements, with a correlation coefficient of 0.96, whilst the ERA‐Interim rainfall products showed significant overestimation in February, March and September. Nevertheless, all remote sensing rainfall products achieved good performance in June and July, showing promising potential to be an alternative to rainfall data for long‐term water resource analysis. Furthermore, ERA‐Interim rainfall products did not perform as well as the other rainfall sources in flow simulations, as evidenced by the large root‐mean‐square error and bias. CPC rainfall products outperformed all other rainfall sources in terms of the Nash–Sutcliffe coefficient at all temporal scales. In addition, the skill scores suggested that the performance of the streamflow simulation improved significantly on a monthly scale, with a higher Nash–Sutcliffe coefficient and smaller total streamflow errors.

Radar rainfall estimates have become increasingly available for hydrological modellers over recent years, especially for flood forecasting and warning over poorly gauged catchments. However, the impact of using radar rainfall as compared with conventional raingauge inputs, with respect to various hydrological model structures, remains unclear and yet to be addressed. In the study presented by this paper, we analysed the flow simulations of the upper Medway catchment of southeast England using the UK NIMROD radar rainfall estimates, using three hydrological models based upon three very different structures (e.g. a physically based distributed MIKE SHE model, a lumped conceptual model PDM and an event-based unit hydrograph model PRTF). We focused on the sensitivity of simulations in relation to the storm types and various rainfall intensities. The uncertainty in radar rainfall estimates, scale effects and extreme rainfall were examined in order to quantify the performance of the radar. We found that radar rainfall estimates were lower than raingauge measurements in high rainfall rates; the resolutions of radar rainfall data had insignificant impact at this catchment scale in the case of evenly distributed rainfall events but was obvious otherwise for high-intensity, localised rainfall events with great spatial heterogeneity. As to hydrological model performance, the distributed model had consistent reliable and good performance on peak simulation with all the rainfall types tested in this study.