Abstract
This study uses machine learning models to investigate the climatic factors that affect monthly rainfall variability in a remote region of Madagascar. By analyzing past weather conditions and relevant climate indices, such as Niño 3.4, Oceanic Niño Index (ONI), Outgoing Long Wave Radiation (OLR), and North Pacific Gyre Oscillation (NPGO), a short- to medium-range rainfall outlook model was developed for the site while integrating the MERRA-2 climate model.
Six statistical rainfall outlook models were constructed to forecast rainfall up to 6 months into the future. A web application was also developed to compile up-to-date rainfall information from station-observed weather records and online climatic gridded models to provide dynamic rainfall outlooks for the area of interest.
These results are intended to be used in an operational water management context with ease of use, enabling more informed decision-making processes for water management projects.
Authors:
Victor Muñoz | SRK Consulting Vancouver, Canada
Lorenzo Bianco | SRK Consulting Cardiff, UK
Mark Sumka | SRK Consulting Vancouver, Canada
Andrew Magee | Rio Tinto Brisbane, Australia
Presenter:
Lorenzo Bianco | SRK Consulting Cardiff, UK
Click here to view other presentations at this event
