Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Diving Deep: Forecasting Sea Surface Temperatures and Anomalies

Ding Ning, Varvara Vetrova, Karin R. Bryan, Yun Sing Koh, Andreas Voskou, N'Dah Jean Kouagou, Arnab Sharma

TL;DR

The methodologies employed, the results obtained, and the lessons learned from the Diving Deep challenge are discussed, offering insights into the future of climate-related predictive modeling.

Abstract

This overview paper details the findings from the Diving Deep: Forecasting Sea Surface Temperatures and Anomalies Challenge at the European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases (ECML PKDD) 2024. The challenge focused on the data-driven predictability of global sea surface temperatures (SSTs), a key factor in climate forecasting, ecosystem management, fisheries management, and climate change monitoring. The challenge involved forecasting SST anomalies (SSTAs) three months in advance using historical data and included a special task of predicting SSTAs nine months ahead for the Baltic Sea. Participants utilized various machine learning approaches to tackle the task, leveraging data from ERA5. This paper discusses the methodologies employed, the results obtained, and the lessons learned, offering insights into the future of climate-related predictive modeling.

Diving Deep: Forecasting Sea Surface Temperatures and Anomalies

TL;DR

The methodologies employed, the results obtained, and the lessons learned from the Diving Deep challenge are discussed, offering insights into the future of climate-related predictive modeling.

Abstract

This overview paper details the findings from the Diving Deep: Forecasting Sea Surface Temperatures and Anomalies Challenge at the European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases (ECML PKDD) 2024. The challenge focused on the data-driven predictability of global sea surface temperatures (SSTs), a key factor in climate forecasting, ecosystem management, fisheries management, and climate change monitoring. The challenge involved forecasting SST anomalies (SSTAs) three months in advance using historical data and included a special task of predicting SSTAs nine months ahead for the Baltic Sea. Participants utilized various machine learning approaches to tackle the task, leveraging data from ERA5. This paper discusses the methodologies employed, the results obtained, and the lessons learned, offering insights into the future of climate-related predictive modeling.
Paper Structure (20 sections, 2 equations, 3 figures)

This paper contains 20 sections, 2 equations, 3 figures.

Table of Contents

  1. Problem Definition
  2. Challenge Setup
  3. Data and Task
  4. Evaluation
  5. Solutions
  6. Solution by Team randomguy
  7. Preliminary analysis
  8. The method
  9. Post Competition Evaluation
  10. Model Performance
  11. Conclusions and Future Work
  12. Solution by Team UPB-DICE
  13. Overview of the Approach.
  14. Main Approach.
  15. Secondary Approach.
  16. ...and 5 more sections

Figures (3)

  • Figure 1: Left: Global average temperature per year, Right: RMSE using N months back values as estimation
  • Figure 2: Comparison of the proposed model with and without correction terms with the persistent model for 2010 to 2023
  • Figure 3: Left: Visualization of model performance on the validation dataset, Right: Predicted vs. target in testing phase