Seminar series | Live-streamed
Enhancing Western United States Sub-Seasonal Forecasts: Forecast Rodeo Prize Competition Series | 1 December | 16:00 GMT
Florian Pappenberger (ECMWF)
Kenneth Nowak is the Water Availability Research Coordinator for the Bureau of Reclamation’s Research and Development Office. In this role, he coordinates internal and external research activities related to streamflow and water supply forecasting, water operations models and decision support systems, open data efforts, and climate variability and change. Prior to joining the Research and Development Office, Ken worked for Reclamation’s Lower Colorado Region on long-term Colorado River planning and management. Ken holds Ph.D. and M.S. degrees in Civil Engineering from the University of Colorado and a B.S. in Environmental Engineering from Rensselaer Polytechnic Institute.
The Bureau of Reclamation (Reclamation) is the largest wholesaler of water in the United States, delivering water to 31 million people, serving 20 percent of western irrigators, and generating enough hydropower to serve 3.5 million homes. Providing these services is informed by a variety of hydroclimate observation and forecast information. Improved sub-seasonal forecasts of temperature and precipitation would enable water managers to better prepare for shifts in hydrologic regimes, such as the onset of drought or occurrence of wet weather extremes.
In support of advancing sub-seasonal prediction, Reclamation launched a series of prize competitions called the Sub-Seasonal Climate Forecast Rodeos. These year-long, real-time forecasting competitions are focused on improving Western U.S. temperature and precipitation for weeks 3 & 4 and weeks 5 & 6. Rodeo I concluded in 2018 with several teams outperforming benchmark forecasts, winning $525,000 in prizes. Based on the success of Rodeo I, and the desire to further improve skill of sub-seasonal forecasts, Rodeo II was launched in 2019 and will conclude fall 2020. Rodeo II features the same year-long competition and forecasts as Rodeo I, but is open to international participation, and winning solutions from Rodeo I serve as benchmarks. Relative to Rodeo I, participation in Rodeo II has increased almost ten-fold, with many teams outperforming the benchmarks.
This presentation will provide an overview of Rodeo I and II, experiences, and final outcomes. Further, it will discuss some of the winning methods for advancing sub-seasonal forecasts of temperature and precipitation in the Western United States.
Recordings and slides of past talks
Causal Networks as a framework for climate science to improve process understanding | 27 October
Speaker: Marlene Kretschmer
Probabilistic downscaling to detect regional present and future climate hazards | 28 April
Speaker: Sherman Lo (University of Warwick)
Exploring Machine Learning for Data Assimilation | 7 May
Speaker: Alban Farchi (ECMWF)
MetNet: A Neural Weather Model for Precipitation Forecasting | 12 May
Speaker: Nal Kalchbrenner (Google Research Amsterdam)
AI, a change in science/technology ... or culture? | 14 July
Speaker: Alberto Arribas (Met Office)
Spatiotemporal complexity and time-dependent networks in mid- to late Holocene simulations | 28 July
Speaker: Dr Annalisa Bracco (Georgia Institute of Technology)
Building trustworthy AI for environmental science | 8 September
Speaker: Amy McGovern (University of Oklahoma)
From research to applications – Examples of operational ensemble post-processing using machine learning |1 October
Speaker: Maxime Taillardat (Météo-France)
Other Machine Learning Talks at ECMWF
Title: "ecPoint” - a Post-processing Tool that improves Forecasts and highlights Systematic Model Errors
Speaker: Timothy Hewson
In April 2019 ECMWF introduced a new, experimental, “point rainfall” forecast product onto its ecCharts web display platform, based on the post-processing package “ecPoint”, to give site-specific forecasts for everywhere in the world up to day 10. Prior to this development forecasters had access to just raw ensemble output in ecCharts, which provides gridbox average totals. ecPoint aims to incorporate probabilistically the expected sub-grid variability, and simultaneously apply gridscale bias corrections. Both these adjustments depend critically on “gridbox-weather-type”.
This presentation will describe the meteorology-based calibration rationale that underpins ecPoint, how this is different to pre-existing post-processing methods, and how it can also be applied to other surface variables such as 2m temperature. Numerous benefits will be highlighted.
The conditional verification concepts underpinning the calibration allow one to identify weather-situation-dependant gridscale biases. Examples will illustrate the diagnostic power of this approach, showing where and when rainfall is typically under- and over-forecast, providing pointers for future model improvements. And using an open source GUI one can apply the calibration code to data from other models, and thereby intercompare performance in different weather situations.
The forecast improvements that then arise will be discussed, using both long term global verification up to day 10, and illustrative case studies, with a focus on how extreme localised rainfall, that might lead to flash floods, is better handled. It will be shown how the post-processing can usefully shift the emphasis for warning issue from one region to another, when one compares with raw ensemble output.
There will be brief reference, from collaborative work, to how ecPoint output seems to compare favourably with the post-processed output of convection-resolving limited area ensembles.
The talk will conclude by discussing, in the context of ongoing and potential projects, numerous future applications of ecPoint, such as bias-corrected inputs to hydrological models, point rainfall re-analyses and tests of theories such as city impact on rainfall. Avenues for improving the methodology will also be highlighted.
The links to join the seminars will be published on this page. If you would like to receive the joining instructions for each seminar by email, please complete the registration form.