Innovative Frameworks to substantiate Accurate and Efficient Drought Monitoring and its Forecasting

Abstract

Drought is a recurring climatic event, which often hits several parts of the world. Its impacts have an enormous effect on the natural environment, ecosystem, society, and many economic and social sectors worldwide. Whether resilient to drought or reducing the drought's negative impacts, understanding the drought's multiple dynamic dimensions (spatial and temporal) is important. Moreover, accurate and precise drought monitoring and prediction techniques strengthen the decision-makers in making critical management decisions related to the drought hazard. Therefore, we develop several dynamic dimensions frameworks to substantiate accurate and efficient drought monitoring, assessment, and forecasting. In this perspective, the thesis presents eight major contributions that allow for a faster and more comprehensive prognosis of drought events and make a possible earlier warning to those affected. In the first contribution of the thesis, we develop a Monte Carlo Feature Selection and Interdependency Discovery (MCFS-ID) algorithm-based framework for accurate estimation and continuous monitoring of drought hazards. The results provide important meteorological stations for further climatological analysis of databases and lead toward minimal use of resources. In the second contribution, we develop a framework that accumulates information from multiple homogenous stations. The proposed framework leads toward minimal use of time and resources for assessing and monitoring drought hazards. The outcomes of the research provide a basis for accurate and comprehensive drought characterization. In the third contribution, we propose a framework to assess spatially accumulated meteorological time series data for the comprehensive characterization of drought hazards in regional drought monitoring. The outcomes associated with this framework support investigators and policymakers to practice the precise and inclusive characterization of drought hazards in presence of influential climatic factors. Our fourth contribution investigates more accurate spatiotemporal characteristics of various drought classes in regional settings. Here, we propose a new Maximum Spatio Temporal Two-Stage Standardized Weighted Index (MSTTSSWI) for regional drought characterization. The outcomes associated with MSTTSSWI help researchers and policymakers more accurately characterize drought hazards. Moreover, we examine seasonal drought frequency and persistence spatiotemporal patterns in the fifth contribution. For this purpose, the logistic regression model is used by describing the DRSML QAU ix spatial pattern of seasonal drought frequency and persistence. The outcomes provide the basis for effective water resource management and agriculture sectors. In the sixth contribution, we develop a Model-Based Clustering of Categorical Drought States Sequences (MBCCDSS), for monthly prediction of drought severity. The outcome of MBCCDSS timely informs decision-makers to bring reliable plans and strategies to decrease the negative impacts of drought occurrences. In the seventh contribution, we develop a Copula-based Modified Continuous Drought Probability Monitoring System (MCDPMS) to promptly monitor drought occurrences. The SSP and most popular Meta Elliptical copulas and Archimedean copulas are used in MCDPMS for evaluating several drought classes. The outcomes associated with MCDPMS provide a new quantifiable method for evaluating and managing drought events. Finally, in the eighth contribution, we present the Regional Spatially Agglomerative Continuous Drought Probability Monitoring System (RSACDPMS) to attain spatiotemporal data for numerous drought classes in a homogenous region. The RSACDPMS is developed to monitor regional drought characteristics more expeditiously. The RSACDPMS may provide an indication of the drought probability and the time-based progression of the drought incidences in a homogenous region. The outcomes of the RSACDPMS allow the users to look forward to improvement and modification approaches, or even to issue alerts