MATHEMATICAL MODELLING FOR DISAGGREGATION OF RAINFALL DATA

Abstract

In many cases the rainfall data at the course-scale is not sufficient for the user, such as farmers and environmental modelers, and they require the rainfall data at a daily scale or a sub daily scale. The requirement of such data is that it helps in the hydrological planning, design and study of the area. This also enables the related agricultural people to make appropriate decisions based on the rainfall use. For them, a model is required which can consistently disaggregate the rainfall data and fulfill their requirement. There are different models present which can be helpful in doing so. The three distinct approaches for disaggregating rainfall sequences, i.e., random cascading model, Bartlett–Lewis model – RBLM, and method of fragments are developed for the purpose. At many places like in Germany, the disaggregation model is used for making the relation between the runoff and the rainfall of the river. This model also helped in the prediction of the flood in the area. A study in Australian cities is done which is used as comparison of the outcomes of the models and develops an intensity duration curve of the areas. Another study in South Africa is done for temporal rainfall disaggregation. These studies have been very effective in the hydrological study of the area and help in development of dams and predicting the flood times dry period spells and wet period spells so that planning can be done accordingly. In this study, the main focus is to analyse the historical data and disaggregate it on a daily and sub daily basis using the random cascade model for the region in Chennai. The historical daily rainfall data is available from the 3 rain gauges for the ten years (1998-2007), whose average is taken to make the list of daily rainfall data. The satellite Page No. 4 data for the same region and the next year is used in the study so as to calculate the fitness of the generated data on the daily bases. The statistics of the disaggregated data are represented and the various deviations from the original data are found out. To understand the disaggregated data, hyetographs are made for different duration intensities. These hyetographs help the targeted users to get the knowledge they require for the particular given time. The hyetographs are also used to determine the dry and the wet spell of the day on a different hourly basis. Everyday wet and dry spells helps in understanding the pattern in which each day the rainfall can vary and the intensity in that duration. These intensity can be utilised for the further study and predictions of the data related to rainfall in fine-scale time units.