Dépôt numérique
RECHERCHER

Lake surface area forecasting using integrated Satellite-SARIMA-Long-Short-Term memory model.

Téléchargements

Téléchargements par mois depuis la dernière année

Plus de statistiques...

Soltani, Keyvan, Azari, Arash, Zeynoddin, Mohammad, Amiri, Afshin, Ebtehaj, Isa, Ouarda, Taha B. M. J. ORCID: https://orcid.org/0000-0002-0969-063X, Gharabaghi, Bahram et Bonakdari, Hossein (9999). Lake surface area forecasting using integrated Satellite-SARIMA-Long-Short-Term memory model. Soft Computing . DOI: 10.21203/rs.3.rs-631247/v1. (Sous Presse)

[thumbnail of P3967.pdf] PDF
Document sous embargo jusqu'à 5 août 2022.

Télécharger (3MB)

Résumé

Lake Water Surface Area (WSA) plays a vital role in environmental preservation and future water resource planning and management. Accurately mapping, monitoring and forecasting Lake WSA changes are of great importance to regulatory agencies. This study used the MODIS satellite images to extract a monthly time series of WSA of two lakes located in Iran from 2001 to 2019. Following a consequence of image and time series preprocessing to obtain the preprocessed lake surface area time series, the outcomes were modeled by the Long-Short-Term Memory (LSTM) deep learning (DL) method, the stochastic Seasonal Auto-Regressive Integrated Moving Average (SARIMA) method and hybridization of these two techniques with the objective of developing WSA forecasts. After separate standardization and normalization of AL TS and reevaluation of the preprocessed data, the SARIMA (1, 0, 0) (0, 1, 1)12 model outperformed sole LSTM models with correlation index of (R) 0.819, mean absolute error (MAE) of 49.425 and mean absolute percentage error (MAPE) of 0.106. On the other hand, the hybridization (stochastic-DL) enhanced the reproduction of the primal statistical properties of WSA data and caused better mediation. However, the other accuracy indices did not change markedly (R 0.819, MAE 49.310, MAPE 0.105). The multi-step preprocessing and reevaluation also caused all LSTM models to produce their best results by less than 12 inputs.

Type de document: Article
Mots-clés libres: water resources; stochastic model; SARIMA; Tashk-Bakhtegan Lakes; hybrid model; forecasting
Centre: Centre Eau Terre Environnement
Date de dépôt: 15 oct. 2021 17:39
Dernière modification: 15 févr. 2022 20:37
URI: https://espace.inrs.ca/id/eprint/11956

Actions (Identification requise)

Modifier la notice Modifier la notice