Machine Learning

Last modified by S2S_mchnlearn on 2020/04/02 15:27

Machine Learning / Artificial Intelligence for S2S prediction 

There is currently a lot excitement in the weather and climate communities to explore the potential of data driven approaches based on Artificial Intelligence/Machine Learning/Deep learning for S2S prediction through, for instance, improved parameterization, improved calibration and multi-model calibration, extreme event attribution, verification... The publicly available S2S database which contains a considerable amount of data (re-forecasts and real-time forecasts from 11 operational centres) represents an ideal testbed for these data-driven methods. The SubX database provides another such opportunity (http://cola.gmu.edu/subx/index.html).

Potential applications :
 1. Improved data assimilation (e.g. better quality control of observations)
 2. Improved parameterization (e.g. radiative schemes)
 3. Improved post-processing (model calibration, bias-correction, multi-ensemble combination...)
 4. Predictability diagnostics (e.g. teleconnections)
 5. S2S event attribution (e.g. origins of extreme events)
 6. Empirical forecasts

http://enterprise.xwiki.org/xwiki/bin/view/UserGuide/

Researches on AI/ML for S2S prediction
1. Scripps Institute of Oceanography  (From Dr. Peter Gibson. Updated April 1 2020)
 The project explores the potential for modern machine learning tools to improve seasonal prediction skill of precipitation over the Western US. Modern machine learning approaches are 'data hungry' while observations are data limited (relatively short in length for the purposes of seasonal forecasting). To circumvent this issue, we train a variety of machine learning tools on perturbed initial condition climate model ensembles that span several thousands of years, then use these 'learnt' teleconnections to make seasonal predictions. We are testing a hierarchy of machine learning approaches from simple to complex: simple logistic regression, LASSO, Random Forests, Gradient Boosted decision trees, and convolutional neural networks. This project is a collaboration between researchers at Scripps CW3E and JPL, and funded by the California Department of Water Resources.  
 

2. Australian Bureau of Meteorology (From Catherine de Burgh-Day​ with Oscar Alves and Debbie Hudson. Updated April 2 2020))

We are at the early stages of work developing a ML-based vegetation model which uses outputs of the Bureau's seasonal prediction ACCESS-S.The purpose of this work is twofold:  

  • Investigate the possibility of making predictions of vegetation ​characteristics in the coming weeks and seasons using model outputs as predictors. Forecasts of vegetation could have potential use for a number of sectors including fire agencies and agriculture 
  • Attempt to use the vegetation model we develop to periodically update the vegetation ancillary file used in model runs. Currently ACCESS-S1 uses a static vegetation file. We plan to investigate what possible skill gains could be got from a more dynamic representation of vegetation, and then to try updating the vegetation ancillary of the model every N timesteps by passing it through our vegetation model, along with the latest model parameters. 

We intend to start by trying an LSTM Neural Network for the vegetation model, potentially also including some convolutional layers. We will however be investigating what is most effective as we go. Initially we will be training using the ACCESS-S1 hindcast, however if a larger training set is needed we may investigate using a larger set to train, followed by transfer learning techniques to update the model to ACCESS-S. 

Tags:
Created by Administrator on 2020/04/01 10:19
    

Applications

This wiki is licensed under a Creative Commons 2.0 license
XWiki Enterprise 6.2.2 - Documentation