Modeling the Climate-Dengue Nexus with XGBoost in Central and South American Nations: Costa Rica

Read full text Summarise with VIRAT
Authors: David Knut Bengtsson-Asgarali
Cited by: 0

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

The working paper file name is "DengX_costa_rica.docx". It can be downloaded below. This paper presents a machine learning analysis of the climate–dengue nexus in Costa Rica using the DengX framework. Suspected dengue case data from PAHO (2013–2025) were combined with climate variables (temperature, relative humidity, precipitation) extracted from ERA5‑Land via Google Earth Engine. XGBoost model approaches were trained and evaluated: a climate‑only model (MA1), a climate plus autoregressive model (MA2), and two baseline models (MA3 constant, MA4 autoregressive‑only). SHAP analysis was used to interpret feature importance and identify climatic risk thresholds. The study provides quantitative thresholds and a reproducible framework for early warning system development and highlights the potential of using Subseasonal-to-Seasonal (S2S) data for diseae forecasting. PNG files included were used to create figures seen in working paper. Csv and xlsx files contain raw and transformed data used to in this study. Txt files contain JSON model data. Access to the DengX computer application is in progress, but I can make it available should there be community interest.

Fields of Study & Subjects

Mosquito-borne diseases and control Climate Change and Health Impacts Dengue and Mosquito Control Research

Publication details

Journal / source
Zenodo (CERN European Organization for Nuclear Research)
Publication date
25 March 2026
Language
EN