The unknown spatial distribution and kinematics of short-term rainfall are recognised as the most important reasons for errors in runoff simulation on urban catchments. Since the storm movement influences the shape of the runoff hydrographs, the modern rainfall input should include patterns of this movement. In order to be able to take rainfall kinematics into account in runoff simulation, rain movement parameters must be extracted from the rainfall data itself or taken from other phenomena such as wind movement at high altitude. The paper presents the results of the application of three different storm tracking methods on a set of rainfall data from the 12-gauge network in Lund, Sweden. The main conclusion is that a really objective and reliable storm tracking method does not exist yet. When more than one rainfall cell is simultaneously present over the rain gauge network, all so-called objective methods fail or must lose their objectivity. On the other hand, when the rainfall data are good, the storm movement pattern can be easily, though subjectively, recognized. Then, simple methods based on triangulation work equally well as more sophisticated methods based on spatial correlation. The existence of a relation between rainfall and high-altitude wind movement parameters is now well documented. However, further studies subjected to derivation of storm movement parameters from gauge data are still important in order to quantify this relationship better and to explain eventual regional differences.
Read full abstract