Filter-wrapper Input Variable Selection for Monthly Runoff Interval Forecasting Using Local and Global Climate Information

DOI

10.2991/978-94-6463-726-7_5How to use a DOI?

Keywords

Monthly runoff forecasting; Wrapper input variable selection; Interval Forecasting; Binary metaheuristic algorithm

Abstract

The use of global climate and local weather information as input variables can effectively enhance the accuracy of data-driven monthly runoff prediction models. However, selecting appropriate inputs from high-dimensional variables remains a challenge in monthly runoff forecasting. Filter-wrapper input variable selection (FIVS-WIVS) can efficiently select input variables (IVS) from high-dimensional data, particularly through the WIVS method, which is based on binary metaheuristic algorithms (BMAs) and is capable of identifying a set of variables that are valuable for prediction. However, the approach has not been systematically studied. Moreover, previous studies have rarely been able to efficiently perform both deterministic and interval forecasting tasks for monthly runoff. Therefore, the study proposed a monthly runoff deterministic and interval forecasting model composed of four BMAs combined with a Gaussian Process Regression model (GPR). The effects of various BMAs on WIVS, as well as their role in forecasting monthly runoff, were examined. And the optimal combination to further enhance interval monthly runoff predictions was studied. The case study focused on the Lhasa River. The results indicated that the wrapper based on GJO-GPR had the most significant effects on dimensionality reduction, yielding accurate monthly runoff prediction, with an NSE value of 0.86 and an MAE of 68.03 m3/s. Meanwhile, the 50% interval forecasting results based on the GJO-GPR wrapper can successfully encompass the observed runoff. The results demonstrated that the proposed GJO-GPR wrapper effectively selects input factors and achieves high runoff prediction accuracy. Additionally, the proposed wrapper can effectively quantify the uncertainty in the monthly runoff forecasting process, demonstrating significant potential for engineering applications.

Copyright

© 2025 The Author(s)

Open Access

Open Access This chapter is licensed under the terms of the Creative Commons Attribution-NonCommercial 4.0 International License (http://creativecommons.org/licenses/by-nc/4.0/), which permits any noncommercial use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

Download article (PDF)

TY  - CONF
AU  - Binlin Yang
AU  - Lu Chen
AU  - Bin Yi
AU  - Siming Li
PY  - 2025
DA  - 2025/06/13
TI  - Filter-wrapper Input Variable Selection for Monthly Runoff Interval Forecasting Using Local and Global Climate Information
BT  - Proceedings of the 2024 6th International Conference on Hydraulic, Civil and Construction Engineering (HCCE 2024)
PB  - Atlantis Press
SP  - 38
EP  - 51
SN  - 2589-4943
UR  - https://doi.org/10.2991/978-94-6463-726-7_5
DO  - 10.2991/978-94-6463-726-7_5
ID  - Yang2025
ER  -