All Issue

2026 Vol.38, Issue 2 Preview Page
Impact of Lookback Window Length on Transformer-Based Multi-Step Forecasting of Chlorophyll-a in an Agricultural Reservoir
Jinyoo Lee1
,
Yeongseo Kim1
,
Jonggun Kim2
,
Seoro Lee3*
1Undergraduate Student, Department of Regional Infrastructure Engineering, Kangwon National University, Chuncheon 24341, Korea
2Associate Professor, Department of Regional Infrastructure Engineering, Kangwon National University, Chuncheon 24341, Korea
3Assistant Professor, Department of Regional Infrastructure Engineering, Kangwon National University, Chuncheon 24341, Korea
*Corresponding Author
30 June 2026. pp. 171-182
PDF XML Citation
Abstract

In this study, a Transformer-based Sequence-to-Sequence (S2S) model for multi-step (72 h) chlorophyll-a (Chl-a) forecasting was developed, and the impact of the lookback window (i.e., historical input sequence length) on the forecasting performance was investigated. By evaluating various input sequence lengths ranging from 12 to 96 h, the results showed that a 48 h lookback window achieved the most robust performance, yielding an average Nash-Sutcliffe Efficiency (NSE) of 0.811 across the entire forecasting horizon and maintaining a stable NSE of 0.762 even at t+72 h. In contrast, the forecasting accuracy at t+72 h declined to NSE values of 0.730 and 0.705 when the 72 h and 96 h input sequences were used, respectively. Cross-attention analysis revealed that the model assigned the highest weights to the most recent past interval (t-1 h－t-7 h). These results suggest that an excessively long lookback window introduces redundant or irrelevant information during training, leading to a degraded forecasting performance. Overall, these findings have highlighted that the optimal configuration of the lookback window is a critical factor for Chl-a forecasting in agricultural reservoirs. This study provides a basis for the design and optimization of future time-series-based water quality forecasting models.

Keywords
Agricultural reservoirs
Lookback window
Multi-step Chl-a forecasting
Transformer-based model
References
1

Akiba, T., Sano, S., Yanase, T., Ohta, T., Koyama, M. (2019) Optuna: A next-generation hyperparameter optimization framework. pp.2623-2631. Proceedings of the 25th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining. ACM, Anchorage, USA.

10.1145/3292500.3330701
2

Amorim, F. d. L. L. d., Rick, J., Lohmann, G., Wiltshire, K. H. (2021) Evaluation of machine learning predictions of a highly resolved time series of chlorophyll-a concentration. Appl Sci 11:7208.

10.3390/app11167208
3

Bieroza, M., Dupas, R., Glendell, M., McGrath, G., Mellander, P. E. (2020) Hydrological and chemical controls on nutrient and contaminant loss to water in agricultural landscapes. Water 12:3379.

10.3390/w12123379
4

Blauw, A. N., Benincà, E., Laane, R. W., Greenwood, N., Huisman, J. (2018) Predictability and environmental drivers of chlorophyll fluctuations vary across different time scales and regions of the North Sea. Prog Oceanogr 161:1-18.

10.1016/j.pocean.2018.01.005
5

Bontempi, G., Ben Taieb, S., Le Borgne, Y. A. (2012) Machine learning strategies for time series forecasting. pp.62-77. In: European big data management and analytics summer school. Springer Berlin Heidelberg, Berlin, Heidelberg, Germany.

10.1007/978-3-642-36318-4_3
6

Busari, I., Sahoo, D., Harmel, R. D., Haggard, B. E. (2023) A review of machine learning models for harmful algal bloom monitoring in freshwater systems. J Nat Resour Agric Ecosyst 1:63-76.

10.13031/jnrae.15647
7

Chen, P., Ye, G., Xu, X., Xi, W., Xu, D. (2024) Water environmental capacity analysis and eutrophication assessment of water-supplied reservoirs. Desalin Water Treat 317:100200.

10.1016/j.dwt.2024.100200
8

Guan, G., Wang, Y., Yang, L., Yue, J., Li, Q., Lin, J., Liu, Q. (2022) Water-quality assessment and pollution-risk early-warning system based on web crawler technology and LSTM. Int J Environ Res Public Health 19:11818.

10.3390/ijerph19181181836142084PMC9517095
9

He, H., Ma, Y. (2013) Imbalanced learning: Foundations, algorithms, and applications. Wiley-IEEE Press, Hoboken, New Jersey, USA.

10.1002/9781118646106
10

Jargal, N., Atique, U., Mamun, M., An, K. G. (2021) Seasonal and long-term connections between trophic status, sestonic chlorophyll, nutrients, organic matter, and monsoon rainfall in a multipurpose reservoir. Water 13:1720.

10.3390/w13131720
11

Jo, Y.-J., Song, J.-H., Her, Y., Provolo, G., Beom, J., Jeung, M., Kim, Y.-J., Yoo, S.-H., Yoon, K.-S. (2021) Assessing the Potential of Agricultural Reservoirs as the Source of Environmental Flow. Water 13:508.

10.3390/w13040508
12

Kim, J., Lee, W. H., Park, J. (2025) Improvement of deep learning model performance for algal bloom prediction by resolving data imbalance in field observations. Environ Earth Sci 84:417.

10.1007/s12665-025-12420-z
13

Nash, J. E., Sutcliffe, J. V. (1970) River flow forecasting through conceptual models part I—A discussion of principles. J Hydrol 10:282-290.

10.1016/0022-1694(70)90255-6
14

Paerl, H. W., Gardner, W. S., Havens, K. E., Joyner, A. R., McCarthy, M. J., Newell, S. E., Qin B., Scott, J. T. (2016) Mitigating cyanobacterial harmful algal blooms in aquatic ecosystems impacted by climate change and anthropogenic nutrients. Harmful Algae 54:213-222.

10.1016/j.hal.2015.09.009
15

Rozemeijer, J., Jordan, P., Hooijboer, A., Kronvang, B., Glendell, M., Hensley, R., Rinke, K., Stutter, M., Bieroza, M., Turner, R., Mellander, P. E., Thorburn, P., Cassidy, R., Appels, J., Ouwerkerk, K., Rode, M. (2025) Best practice in high-frequency water quality monitoring for improved management and assessment; A novel decision workflow. Environ Monit Assess 197:353.

10.1007/s10661-025-13795-z40038155PMC11880092
16

Vaswani, A., Shazeer, N., Parmar, N., Uszkoreit, J., Jones, L., Gomez, A. N., Kaiser, Ł., Polosukhin, I. (2017) Attention is all you need. Advances in Neural Information Processing Systems. NeurIPS, Long Beach, USA.

17

Wen, Q., Zhou, T., Zhang, C., Chen, W., Ma, Z., Yan, J., Sun, L. (2023) Transformers in time series: A survey. pp.6778-6786. Proceedings of the 32nd International Joint Conference on Artificial Intelligence. IJCAI, Macao, China.

10.24963/ijcai.2023/759
18

Zahir, M., Su, Y., Chen, Y., Shahzad, M. I., Ayub, G., Rahman, S. U., Ahmed, T., Ijaz, J. (2025) Anthropogenic and climate‐driven changes on harmful algal blooms in two chinese reservoirs. Ecohydrology 18:e2745.

10.1002/eco.2745
19

Zamani, M. G., Nikoo, M. R., Jahanshahi, S., Barzegar, R., Meydani, A. (2023) Forecasting water quality variable using deep learning and weighted averaging ensemble models. Environ Sci Pollut Res 30:124316-124340.

10.1007/s11356-023-30774-4
20

Zhang, S., Arhonditsis, G. B., Ji, Y., Bryan, B. A., Peng, J., Zhang, Y., Gao, J., Zhang, J., Cho, K. H., Huang, J. (2025) Climate change promotes harmful algal blooms in China’s lakes and reservoirs despite significant nutrient control efforts. Water Res 277:123307.

10.1016/j.watres.2025.123307
21

Zhou, H., Zhang, S., Peng, J., Zhang, S., Li, J., Xiong, H., Zhang, W. (2021) Informer: beyond efficient transformer for long sequence time-series forecasting. pp.11106-11115. Proceedings of the AAAI Conference on Artificial Intelligence. AAAI, Palo Alto, USA.

10.1609/aaai.v35i12.17325
Information
  • Publisher :Agriculture and Life Sciences Research Institute, Kangwon National University
  • Publisher(Ko) :None
  • Journal Title :Journal of Agricultural, Life and Environmental Sciences
  • Volume : 38
  • No :2
  • Pages :171-182
  • Received Date : 2026-04-21
  • Revised Date : 2026-04-28
  • Accepted Date : 2026-04-28
  • DOI :https://doi.org/10.22698/jales.20260012
Journal Informaiton Journal of Agricultural, Life and Environmental Sciences Journal of Agricultural, Life and Environmental Sciences
Online Submission & Review System
  • NRF
  • KOFST
  • KISTI Current Status
  • KISTI Cited-by
  • KCI 문헌 유사도 검사_영문_불교학연구
  • orcid
  • open access
  • ccl
Journal Informaiton Journal Informaiton - close