Models Forecasting and Evaluation of Karachi Rainfall
Main Article Content
Abstract
Building up a model for hydrological forecasting from historical records is essential to powerful hydropower reservoir management, implementation, controlling, and discharging water. Conventionally, modelling and analysis through a time series approach have been used to build mathematical models for producing hydrologic records in water resources and hydrology. Recently artificial intelligence (AI) is used for forecasting and modelling hydrological data. By using AI, large-scale and long-series data can be analysed with reasonable accuracy. This paper is focused on assessing and forecasting rainfall in Karachi. For this purpose, the daily rainfall data are considered from the period 2006 to 2017. Three novel models Autoregressive Moving Average (ARMA), Seasonal Autoregressive Integrated Moving Average (SARIMA), and Adaptive Neuro-Fuzzy Inference System (ANFIS) are employed. The statistical measure; Root Mean Squared Error (RMSE) is used to evaluate the performance of these models. We have also determined the values for the Akaike information criterion (AIC), Bayesian Schwarz information criterion (SIC), Hannan Quinn information criterion (HIC), and Durbin-Watson test (DW), Mean Absolute Error (MAE), and Mean Absolute Percentage Error (MAPE). Estimation of these parameters assisted in selecting the best-fitted model. Findings indicate that ANFIS and SARIMA outperform contrary to all evaluation criteria. This study implies researchers embrace ANFIS and SARIMA for predicting rainfall, hydrology, and water resources with high accuracy.
Article Details

This work is licensed under a Creative Commons Attribution 4.0 International License.
Papers accepted for publication become copyright of the Editorial Office, Journal of Mountain Area Research, Karakoram International University Gilgit, Gilgit-15100, Pakistan and authors will be asked to sign a copyright transfer agreement. Articles cannot be published until a signed copyright transfer agreement form has been received.
The work published in this journal is licensed under a Creative Commons Attribution 4.0 International License.