Incorporation of weather parameters in MMRC-K model for rainfall disaggregation

Rainfall disaggregation techniques are employed to convert low-resolution rainfall data into high-resolution rainfall series when high-resolution data is not readily available, enabling a more detailed analysis of rainfall patterns and intensities. The Microcanonical Multiplicative Random Cascade (MMRC) model is a commonly used stochastic method for disaggregating rainfall data into finer temporal or spatial scales. The traditional MMRC model overestimates extreme rainfall values due to its reliance on predefined classes. To mitigate this issue, clustering techniques are employed to classify rainfall data based on similar characteristics. The integration of K-means clustering into the traditional MMRC model has led to notable enhancements in capturing extreme rainfall characteristics. Temperature, wind speed, and humidity, among other factors, play crucial roles in influencing rainfall patterns. By incorporating these data into the clustering process, this study aims to classify rainfall more accurately, resulting in improved preservation of extreme rainfall characteristics. The integration of additional weather parameters into K-means clustering enhanced the statistical properties of the rainfall series generated for German cities. In contrast, the IDF curves produced by this model for the Indian cities exhibit greater reliability compared to those generated by the conventional MMRC model. However, the comparison of the modified MMRC-K models using the weather parameters with the MMRC-K model, did not show improvement in preserving extreme rainfall characteristics. Since the dataset required for this modified model is more extensive, the MMRC-K model becomes the more suitable option.