Prediction of Sea Level with Vertical Land Movement Correction Using Deep Learning

Nawin Raj ()
Additional contact information
Nawin Raj: School of Mathematics, Physics and Computing, Springfield Campus, University of Southern Queensland, Toowoomba, QLD 4300, Australia

Mathematics, 2022, vol. 10, issue 23, 1-23

Abstract: Sea level rise (SLR) in small island countries such as Kiribati and Tuvalu have been a significant issue for decades. There is an urgent need for more accurate and reliable scientific information regarding SLR and its trend and for more informed decision making. This study uses the tide gauge (TG) dataset obtained from locations in Betio, Kiribati and Funafuti, Tuvalu with sea level corrections for vertical land movement (VLM) at these locations from the data obtained by the Global Navigation Satellite System (GNSS) before the sea level trend and rise predictions. The oceanic feature inputs of water temperature, barometric pressure, wind speed, wind gust, wind direction, air temperature, and three significant lags of sea level are considered in this study for data modeling. A new data decomposition method, namely, successive variational mode decomposition (SVMD), is employed to extract intrinsic modes of each feature that are processed for selection by the Boruta random optimizer (BRO). The study develops a deep learning model, namely, stacked bidirectional long short-term memory (BiLSTM), to make sea level (target variable) predictions that are benchmarked by three other AI models adaptive boosting regressor (AdaBoost), support vector regression (SVR), and multilinear regression (MLR). With a comprehensive evaluation of performance metrics, stacked BiLSTM attains superior results of 0.994207, 0.994079, 0.988219, and 0.899868 for correlation coefficient, Wilmott’s Index, the Nash–Sutcliffe Index, and the Legates–McCabe Index, respectively, for Kiribati, and with values of 0.996806, 0.996272, 0.992316, and 0.919732 for correlation coefficient, Wilmott’s Index, the Nash–Sutcliffe Index, and the Legates–McCabe Index, respectively, for the case of Tuvalu. It also shows the lowest error metrics in prediction for both study locations. Finally, trend analysis and linear projection are provided with the GNSS-VLM-corrected sea level average for the period 2001 to 2040. The analysis shows an average sea level rate rise of 2.1 mm/yr for Kiribati and 3.9 mm/yr for Tuvalu. It is estimated that Kiribati and Tuvalu will have a rise of 80 mm and 150 mm, respectively, by the year 2040 if estimated from year 2001 with the current trend.

Keywords: sea level rise (SLR); Global Navigation Satellite System (GNSS); vertical land movement (VLM); successive variational mode decomposition (SVMD); bidirectional long short-term memory (BiLSTM); Boruta random forest optimizer (BRFO) (search for similar items in EconPapers)
JEL-codes: C (search for similar items in EconPapers)
Date: 2022
References: View references in EconPapers View complete reference list from CitEc
Citations: View citations in EconPapers (1)

Downloads: (external link)
https://www.mdpi.com/2227-7390/10/23/4533/pdf (application/pdf)
https://www.mdpi.com/2227-7390/10/23/4533/ (text/html)

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:gam:jmathe:v:10:y:2022:i:23:p:4533-:d:989381

Access Statistics for this article

Mathematics is currently edited by Ms. Emma He

More articles in Mathematics from MDPI
Bibliographic data for series maintained by MDPI Indexing Manager ().