期刊
ENVIRONMENT DEVELOPMENT AND SUSTAINABILITY
卷 23, 期 2, 页码 1392-1410出版社
SPRINGER
DOI: 10.1007/s10668-020-00626-z
关键词
ARIMA; Forecasting; Radiometric water indices; S-ARIMA; Seasonality
资金
- Deanship of Scientific Research (DSR) at King Abdulaziz University, Jeddah [KEP-MSc-01-155-38]
Monitoring water quality parameters of inland water resources in arid environments is crucial for their control and management. Results suggest that the S-ARIMA model is more reliable than the ARIMA model for predicting water quality parameters, especially in scenarios with seasonal features.
The monitoring of inland water resources in arid environments is an essential element due to their fragility. Reliable prediction of the water quality parameters helps to control and manage the water resources in arid regions. Water quality parameters were estimated using remote sensing data acquired from the beginning of 2017 until the end of 2018. The prediction of the water quality parameters was comprehended by using an adjusted autoregressive integrated moving average (ARIMA) and its extension seasonal ARIMA (S-ARIMA). Maximum Chlorophyll Index (MCI), Green Normalized Difference Vegetation Index (GNDVI) and Normalized Difference Turbidity Index (NDTI) were the tested water quality parameters using Sentinel-2 sensor on temporal resolution basis of the sensor. Results indicated that the implementation of the ARIMA model failed to sustain a reliable prediction longer than one-month time while S-ARIMA succeeded to maintain a robust prediction for the first 3 months with confidence level of 96%. MCI has its ARIMA at (1,2,2) and S-ARIMA at (1,2,2) (2,1,1)6, GNDVI has its ARIMA at (2,1,2) and S-ARIMA at (2,1,2) (2,2,2)6, and finally, NDTI has its ARIMA at (2,2,2) and S-ARIMA at (2,2,2) (1,1,2)6. The accuracy of S-ARIMA predictions reached 82% at 6-month prediction period. Meanwhile, there was no solid prediction model that lasted till 12 months. Each of the forecasted water quality parameters is unique in its prediction settings. S-ARIMA model is a more reliable model because the seasonality feature is inherited within the forecasted water quality parameters.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据