325
edits
Seasonal forecasting: Difference between revisions
→Subseasonal Forecasting
(extended the application of ML to seasonal forecasting by introducing the challenge of polar vortex prediction. It may fall more into the extreme weather event prediction.) |
|||
| (8 intermediate revisions by 2 users not shown) | |||
Line 6:
==Background Readings==
=== Subseasonal Forecasting ===
Subseasonal forecasting is the task of predicting the climate of a region between 2-8 weeks in advance. Weather and seasonal prediction which focus on forecasting climate 1-7 days and 2+ months in advance respectively have already received significant attention and are considered easier prediction problems than the subseasonal scenario <ref name=":0">{{Cite journal|last=He|first=Sijie|last2=Li|first2=Xinyan|last3=DelSole|first3=Timothy|last4=Ravikumar|first4=Pradeep|last5=Banerjee|first5=Arindam|date=2020-06-24|title=Sub-Seasonal Climate Forecasting via Machine Learning: Challenges, Analysis, and Advances|url=http://arxiv.org/abs/2006.07972|journal=arXiv:2006.07972 [cs, stat]}}</ref>. Improvements in subseasonal prediction will be realized in industries such as water management <ref name="hwang2019improving">{{Cite journal|last=Hwang|first=Jessica|last2=Orenstein|first2=Paulo|last3=Cohen|first3=Judah|last4=Pfeiffer|first4=Karl|last5=Mackey|first5=Lester|date=2019-07-25|title=Improving Subseasonal Forecasting in the Western U.S. with Machine Learning|url=https://doi.org/10.1145/3292500.3330674|journal=Proceedings of the 25th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining|series=KDD '19|location=Anchorage, AK, USA|publisher=Association for Computing Machinery|pages=2325–2335|doi=10.1145/3292500.3330674|isbn=978-1-4503-6201-6}}</ref>, agricultural productivity, and emergency planning for extreme weather events <ref name=":0" />.
*'''"Sub-Seasonal Climate Forecasting via Machine Learning: Challenges, Analysis, and Advances" (2020) '''<ref name=":0" />: A review showing that machine learning models can be applied generally to the subseasonal forecasting problem context and they highlight the potential for tailored models to make large improvements over existing methods.
* '''"Improving Subseasonal Forecasting in the Western U.S. with Machine Learning" (2019) '''<ref name="hwang2019improving />: The paper develops and applies two distinct nonlinear regression models to the western United States, both of which outperform the baseline model significantly.
* '''"Sub-seasonal forecasting with a large ensemble of deep-learning weather prediction models" (2021) '''<ref name="weyn2021subseasonal">{{Cite journal|last=Weyn|first=Jonathan A.|last2=Durran|first2=Dale R.|last3=Caruana|first3=Rich|last4=Cresswell-Clay|first4=Nathaniel|date=2021-02-09|title=Sub-seasonal forecasting with a large ensemble of deep-learning weather prediction models|url=http://arxiv.org/abs/2102.05107|journal=arXiv:2102.05107 [physics]}}</ref>: The paper proposes a deep learning ensemble model that competes with, but does not outperform existing methods (i.e. the model in use by the European Centre for Medium-Range Weather Forecasts, [https://en.wikipedia.org/wiki/European_Centre_for_Medium-Range_Weather_Forecasts ECMWF]). The proposed model is much more computationally efficient and this result suggests that the authors' research trajectory holds promise for surpassing the conventional methods.
==Conferences, Journals, and Professional Organizations==
Line 12 ⟶ 18:
==Data==
* [https://dataverse.harvard.edu/dataset.xhtml?persistentId=doi:10.7910/DVN/IHBANG SubseasonalRodeo:] A benchmark dataset consisting of 12 features (including temperature, precipitation, humidity, etc.) from 14 different data sources used for training and evaluating subseasonal forecast systems in the contiguous western United States<ref>{{Cite journal|last=Hwang|first=Jessica|last2=Orenstein|first2=Paulo|last3=Cohen|first3=Judah|last4=Mackey|first4=Lester|date=2019-09-24|title=The SubseasonalRodeo Dataset|url=https://dataverse.harvard.edu/dataset.xhtml?persistentId=doi:10.7910/DVN/IHBANG|language=en|doi=10.7910/DVN/IHBANG}}</ref>.
==Future Directions==
| |||