Hybridized artificial intelligence models with nature-inspired algorithms for river flow modeling: A comprehensive review, assessment, and possible future research directions

TY - JOUR

T1 - Hybridized artificial intelligence models with nature-inspired algorithms for river flow modeling

T2 - A comprehensive review, assessment, and possible future research directions

AU - Tao, Hai

AU - Abba, Sani I.

AU - Al-Areeq, Ahmed M.

AU - Tangang, Fredolin

AU - Samantaray, Sandeep

AU - Sahoo, Abinash

AU - Siqueira, Hugo Valadares

AU - Maroufpoor, Saman

AU - Demir, Vahdettin

AU - Dhanraj Bokde, Neeraj

AU - Goliatt, Leonardo

AU - Jamei, Mehdi

AU - Ahmadianfar, Iman

AU - Bhagat, Suraj Kumar

AU - Halder, Bijay

AU - Guo, Tianli

AU - Helman, Daniel S.

AU - Ali, Mumtaz

AU - Sattar, Sabaa

AU - Al-Khafaji, Zainab

AU - Shahid, Shamsuddin

AU - Yaseen, Zaher Mundher

PY - 2024/3

Y1 - 2024/3

N2 - River flow (Qflow) is a hydrological process that considerably impacts the management and sustainability of water resources. The literature has shown great potential for nature-inspired optimized algorithms (NIOAs), like hybrid artificial intelligence (HAI) models, for Qflow modeling. Qflow forecasting needs to be accurate, robust, reliable, and capable of resolving complex non-linear problems to support the decision authority in local and national governments and NGOs. This extensive survey provides a literature review of 100-plus high-impact factor journal articles on developing NIOAs models during 2000–2022. This encompasses a comprehensive review of the established research in different climatic zones, NIOA types, artificial intelligence (AI) models, the input parameters used for model development, Qflow on different time scales, and model evaluation using a wide range of performance metrics. The review also assessed and evaluated several components of relevant literature, along with detailing the existing research gaps. Moreover, the global research gap with future direction is discussed based on current research limitations and possibilities. The data availability evaluation and futuristic suggestions are drafted logically. The review revealed the superiority of the NIOAs among all applied algorithms in the literature. Further, the review concludes that there is a need to improve technical aspects of Qflow forecasting and bridge the gap between scientific research, hydrometeorological model development, and real-world flood and drought management using probabilistic nature inspired (NI) forecasts, especially through effective communication.

AB - River flow (Qflow) is a hydrological process that considerably impacts the management and sustainability of water resources. The literature has shown great potential for nature-inspired optimized algorithms (NIOAs), like hybrid artificial intelligence (HAI) models, for Qflow modeling. Qflow forecasting needs to be accurate, robust, reliable, and capable of resolving complex non-linear problems to support the decision authority in local and national governments and NGOs. This extensive survey provides a literature review of 100-plus high-impact factor journal articles on developing NIOAs models during 2000–2022. This encompasses a comprehensive review of the established research in different climatic zones, NIOA types, artificial intelligence (AI) models, the input parameters used for model development, Qflow on different time scales, and model evaluation using a wide range of performance metrics. The review also assessed and evaluated several components of relevant literature, along with detailing the existing research gaps. Moreover, the global research gap with future direction is discussed based on current research limitations and possibilities. The data availability evaluation and futuristic suggestions are drafted logically. The review revealed the superiority of the NIOAs among all applied algorithms in the literature. Further, the review concludes that there is a need to improve technical aspects of Qflow forecasting and bridge the gap between scientific research, hydrometeorological model development, and real-world flood and drought management using probabilistic nature inspired (NI) forecasts, especially through effective communication.

KW - Data availability

KW - Machine learning

KW - Nature-inspired algorithms

KW - Optimization algorithms

KW - River flow modeling

UR - http://www.scopus.com/inward/record.url?scp=85178647877&partnerID=8YFLogxK

U2 - 10.1016/j.engappai.2023.107559

DO - 10.1016/j.engappai.2023.107559

M3 - Journal article

AN - SCOPUS:85178647877

SN - 0952-1976

VL - 129

JO - Engineering Applications of Artificial Intelligence

JF - Engineering Applications of Artificial Intelligence

M1 - 107559

ER -