تحلیل ویژگی‌های خشکسالی با استفاده از روش مفصل و ترکیب الگوریتم های محاسبات نرم (مطالعه موردی: ایستگاه قزوین)

خشکسالی یک پدیده طبیعی است که تاثیرات منفی قابل توجهی بر بخش‌های مختلف دارد، به‌ویژه به دلیل تغییرات در الگوهای بارندگی مانند مقدار، شدت و توزیع که منجر به خشکسالی‌های سالانه در برخی مناطق شده است. پیش‌بینی، پایش و مدیریت خشکسالی به‌عنوان پدیده‌ای طبیعی همواره از چالش‌های مهم جامعه‌ علمی و تصمیم‌گیری بوده است. برای درک بهتر و پیش‌بینی خشکسالی، پژوهشگران معمولاً از روش‌های آماری در مطالعات هیدرولوژیکی استفاده می‌کنند. توابع احتمال مشترک برای تحلیل چندین متغیر به‌طور همزمان و حفظ همبستگی‌های آن‌ها موثر هستند. علاوه بر این، استفاده از الگوریتم‌های محاسبات نرم برای پیش‌بینی و برآورد به‌طور فزاینده‌ای محبوب شده است. بخش‌های مختلف کشور نیز با توجه به ویژگی‌های جغرافیایی با شدت و مدت متفاوتی از خشکسالی مواجه بوده است. در این مقاله به‌منظور پایش خشکسالی‌های هواشناسی ایستگاه قزوین در ایران، مدت و شدت خشکسالی با استفاده از داده‌های بارندگی ماهانه طی سال‌های 1343 تا 1397 با استفاده از الگوریتم‌های ترکیبی محاسبات نرمm5-pso، m5-ga، m5-gmdh ،m5-gep  وcart بر مبنای توابع مفصل با نرم‌افزار کدنویسی r و genexprotools 4 مورد بررسی قرار گرفته است. ویژگی‌های خشکسالی از شاخص استاندارد بارش ماهانه (spi) محاسبه می‌شود. الگوریتم درخت m5 از نوع دسته‌بندی برای داده‌ها استفاده شد. داده‌های پیش‌بینی باران‌سنجی (از سال 1397-1401) به‌طور تصادفی که 70 درصد داده‌ها برای آموزش و 30 درصد داده‌ها برای آزمون برگزیده شد و سپس با خروجی پژوهش‌های دیگر (میزان خطا، همبستگی و صحت) مقایسه شدند. در بین الگوریتم m5-gep بهترین جواب در مقایسه با سایر مدل‌ها نمایش داد. مقدار rmse برای m5-gep مبتنی بر مفصل 0.12 به‌دست آمد و برتری این مدل را در مقایسه با سایر الگوریتم‌های اشاره شده نشان‌می‌دهد. چنین ویژگی‌های احتمالی خشکسالی می‌تواند برای مدیریت و برنامه‌ریزی منابع آب استفاده شود.

analysis of drought characteristics using a copula-based method and a hybrid of soft computing algorithms (case study: qazvin station)

introduction drought is a natural phenomenon that has significant negative impacts on various sectors, particularly due to changes in rainfall patterns—such as quantity, intensity, and distribution—which have led to annual droughts in certain regions. climate change further complicates this issue by contributing to rising sea levels and altered weather conditions. the gradual onset of drought affects critical areas like water resources, agriculture, and the environment, yet the complex physical factors involved are not well understood, and reliable models are lacking. to better understand and predict drought, researchers often employ statistical methods in hydrological studies. joint probability functions are effective for analyzing multiple variables while preserving their correlations. additionally, the use of soft computing algorithms has become increasingly popular for prediction and estimation. over the past two decades, advancements in data mining techniques have enabled researchers to extract valuable insights from large datasets and create predictive models. these methods are advantageous as they offer greater accuracy and efficiency compared to traditional numerical modeling without incurring laboratory costs. this study specifically examines drought occurrences in qazvin province by integrating joint probability functions and soft computing algorithms to analyze random data. the goal is to improve understanding and management of drought impacts in the region, thereby contributing to more effective responses to this pressing issue. materials and methods this article analyzes monthly rainfall data from january 1964 to december 2018 at the qazvin synoptic station to identify drought characteristics in qazvin province. the average annual rainfall recorded at the station is 304 millimeters. the study employs a combination of copula functions and algorithms to model meteorological drought, utilizing randomly selected data for analysis. specifically, the m5 algorithm is used to generate rainfall data, while various soft computing techniques—including particle swarm optimization (pso), genetic algorithms, cart, gep, and gmdh algorithms—are applied. for drought analysis, the standardized precipitation index (spi) is utilized, as it is recognized as an effective measure for assessing drought conditions. the spi time series derived from monthly rainfall data helps define and calculate key drought characteristics such as duration and severity. drought is identified when the spi falls below -1, with moist and dry conditions classified accordingly. notably, some drought events may exhibit long durations without consistently reaching an spi of -1 or lower; however, cumulative spi values can still indicate drought conditions. a drought event is characterized by a continuous period of negative spi values, with drought duration defined as the length of this negative period and severity measured by cumulative spi during the drought. this research aims to enhance understanding of drought dynamics in qazvin province for better management strategies. results and discussion the amount of rainfall at the qazvin station has been estimated and predicted using the mentioned algorithms and copula forecasting and estimation methods. for the application of joint functions and analysis of two-variable drought, the correlation between the variables should first be estimated. the comparison between observed drought variables (i.e., duration and intensity) and their corresponding fitted distributions indicates a significant correlation between the fitting distribution and the observed drought data. modeling was performed using soft computing algorithms and rainfall data (70% for testing and 30% for validation) and the output of the joint functions to estimate and predict the correlation and error rates with m5-pso, m5-ga, cart, m5-gep, and m5-gmdh algorithms. based on the modeling performance using soft computing and statistical indicators, the m5-gep model is the best. conclusion in this article, the analysis of meteorological drought characteristics (i.e., duration and intensity) for the qazvin station during the years 1964 to 2018 was examined using a combination of the piecewise function with soft computing algorithms. the m5 model was used for parameter classification. one of the important steps in using the combination of algorithms with the piecewise function for analyzing hydrological issues is the appropriate estimation of the dependent parameter of the piecewise function. therefore, inaccurate estimation of the piecewise function parameter leads to underlying or overestimation in the modeling method, which was the main reason for our use of algorithms. the results showed that the m5-gep model could be successfully used for drought modeling. the error rate and correlation coefficient of the model were rmse=0.12 and cc=0.96, respectively, indicating that the m5-gep model provided the best result among the models. additionally, the cart model yielded an error rate and correlation coefficient of rmse=0.46 and cc=0.87, respectively, for estimating drought duration and intensity, which was the worst result among the proposed models.