Evaluation of Meteorological Signals for Drought Forecasting, Using Regression Methods and Artificial Neural Networks
Publish place: 1st Iran-Korea Joint Workshop on Climate Modeling
Publish Year: 1384
نوع سند: مقاله کنفرانسی
زبان: English
View: 1,865
This Paper With 12 Page And PDF Format Ready To Download
- Certificate
- من نویسنده این مقاله هستم
استخراج به نرم افزارهای پژوهشی:
شناسه ملی سند علمی:
IKWCM01_022
تاریخ نمایه سازی: 22 مرداد 1385
Abstract:
Drought is one of the destructive natural disasters, which causes most damages to water resources. Drought forecasting can playa crucial role in water management and optimum operation of water resources. In this research work, it was tried to forecast one year ahead drought status with aid the Arterial Nerrral Networks (ANNs) technique and time series of the SPI and ED1 drought indices. In addition to the indices; rainfalls and large scale meteorological index (i.e. SO1 and NAG) were introduced to the ANNE) as inputs that were not effective as the drought indices. The results showed that the selected algorithm was able to forecast the coming six months drought or wet classes correctly in 80% of the months. These amounts for the nine months ahead were 68% and 60% and for the twelve months are 63 % and 58%, which are eenerally considered to be close results. Finally, comparison of the results for the two indices u revealed that the eITors are more frequent in case of the SPI, such that up to 3 and 4 class difference between the forecasted and the observed ones W8,S observed, while for the ED1 it was Jess than 2 classes.
Keywords:
Authors
Kiarash Bagherzadeh
Tarbiat modarres University, College of Agriculture, Iran
Saeid Morid
Tarbiat modarres University, College of Agriculture, Iran, Corresponding author
Ghaemi
Iran Meteorologycal Organization , Tehran , Iran
مراجع و منابع این Paper:
لیست زیر مراجع و منابع استفاده شده در این Paper را نمایش می دهد. این مراجع به صورت کاملا ماشینی و بر اساس هوش مصنوعی استخراج شده اند و لذا ممکن است دارای اشکالاتی باشند که به مرور زمان دقت استخراج این محتوا افزایش می یابد. مراجعی که مقالات مربوط به آنها در سیویلیکا نمایه شده و پیدا شده اند، به خود Paper لینک شده اند :