Research to Operations (R2O) and S2S forecast and verification products development

Last modified by s2s_wiki on 2020/06/16 06:49

Membership

Caio Coelho (CPTEC/INPE, Brazil)
Andrew Robertson (IRI, USA)
Arun Kumar (NOAA, USA)
Yuhei Takaya (JMA, Japan)
Anca Brookshaw (ECMWF)
Debra Hudson (BoM, Australia)
Angel Muñoz (IRI, USA)
Joanne Robbins (UKMO, UK)

1) Scientific and Operational Objectives

  • Pursue research for testing and developing methodologies for calibration, multi-model combination, verification and generation of forecast products.
  • Coordinate with the relevant WMO technical commissions to define the standards and protocols for operational implementation and exchange of S2S forecasts such that by the end of the Phase II of the S2S, the infrastructure related to the data exchange to support research can be transitioned into the operational domain.

The R2O and S2S forecast and verification products activities plan provides additional information about this sub-project, which builds on the work of the previous (Phase I) S2S Verification and Products sub-project.

2) Linkages with WMO activities

On the research side this sub-project has linkages with the Joint Working Group on Forecast Verification Research (JWGFVR), a WMO joint working group of the Working Group on Numerical Experimentation (WGNE) and the World Weather Research Programme (WWRP). This sub-project also has synergies with the World Climate Research Programme (WCRP) Working Group on Subseasonal to Interdecadal Prediction (WGSIP) through the Climate forecast information for decision making (I4D) project.

On the operational side, this sub-project has linkages with Inter-Programme Expert Team on Operational Predictions from Sub-seasonal to Longer-Time Scale (IPET-OPSLS), a WMO joint team of the Commission for Basic Systems (CBS) and the Commission for Climatology (CCl).

3) Proposed questions to be addressed

The World Weather Research Programme (WWRP) has flagged improving forecasts of precipitation over land as an important area for S2S to focus research and services development efforts. In order to help advance scientific knowledge and the development of forecast and verification products in this priority area this sub-project invites the S2S research and operational communities to address the following questions:

  • What is the current performance level of sub-seasonal precipitation forecasts over land? Over which continental regions can these forecasts be best trusted? How performance levels vary through the seasons of the year?

  • What is the current capability of S2S models in anticipating the occurrence of extreme precipitation events over land (periods of deficit or excess precipitation)?

  • How well the main patterns of precipitation variability on the sub-seasonal time scale over various continental regions are represented in S2S prediction models?

  • How best to combine and calibrate sub-seasonal precipitation forecasts over land in order to produce improved, combined and well-calibrated products and services?

  • Are there identifiable opportunities for producing sub-seasonal precipitation forecasts over land with improved quality? For example, are forecasts produced during Madden and Julian Oscillation (MJO) and/or El Niño Southern Oscillation (ENSO) events more skilful than when neutral conditions are present? Are forecasts for active and break rainfall phases and dry/wet spells (or other quantities of interest) of adequate quality for developing forecast products for use in application sectors?

In order to address these questions the research and operational communities are encouraged to explore existing and develop novel methodologies for forecast calibration, combination and verification. Following the S2S verification chapter produced by the JWGFVR for the recent S2S book, it is particularly encouraged the identification of the most relevant forecast quality attributes for the target audiences (e.g. model and forecast developers, and various application sectors) in order to choose appropriate scores and metrics to be able to adequately address clearly and previously defined verification questions of interest. This practice helps performing a thorough assessment of sub-seasonal forecasts from both the probabilistic and deterministic points of view.

4) Current work of S2S operational and research communities on calibration, multi-model combination, verification and forecast products generation

4.1) Pilot real-time sub-seasonal multi-model ensemble (MME) forecasts and verification products at the WMO LC-LRFMME

The WMO Lead Centre for Long Range Forecast Multi-Model Ensemble (LC-LRFMME) has been developing a pilot system for real-time multi-model subseasonal forecasts using real-time forecasts (and hindcasts) from a subset of models contributing to the WWRP/WCRP S2S research project accessible via ECMWF data archive. Following this link the S2S research and operational communities have the opportunity to see the characteristics of the pilot real-time sub-seasonal MME prediction system developed by the LC-LRFMME, which includes forecast and verification products. These additional slides provide examples of products developed in support of future global, regional and national operational activities performed by WMO members. Subseasonal models from eight Global Producing Centers (GPCs) are currently used: Beijing, ECMWF, Exeter, Melbourne, Montreal, Seuol, Tokyo and Washington. A range of forecast products has been developed including probabilities for tercile categories of weekly/fortnightly averages of 2m temperature and precipitation as well as the MJO and BSISO indices. Hindcast verification has also been generated using ROC curves and scores, reliability diagrams, root mean square error and correlation between hindcast and observed anomalies.

4.2) IRI initiative on investigating the seasonality of subseasonal rainfall and temperature global prediction skill

This document summarizes the methods followed to conduct a global predictive skill assessment for uncalibrated rainfall and 2-m temperature forecasts, produced using the ECMWF’s IFS model, available through the WWRP/WCRP S2S Prediction Project Database via the International Research Institute for Climate and Society (IRI) Data Library.

4.3) Projects and networks dealing with S2S predictions

ACToday: The Adapting Agriculture to Climate Today, for Tomorrow project.

African SWIFT: Science for Weather Information and Forecasting Techniques.

CLIMAX: Climate Services Through Knowledge Co-Production: A Euro-South American Initiative for Strengthening Societal Adaptation Response to Extreme Events.

CSSP Brazil: Climate Science for Service Partnership Brazil

CSSP China: Climate Science for Service Partnership China

SNAP: Stratospheric Network for the Assessment of Predictability. See verification plans on page 6 of this presentation.

WCSSP India: Weather and Climate Science for Service Partnership India

4.4) Software tools

PyCPT of the International Research Institute for Climate and Society (IRI)

4.5) Web portals

IRI Sub-seasonal forecasts map room

CLIMAX project forecast maps: Precipitation, 2 meter temperature, 200 hPa Geopotential height and Outgoing Longwave Radiation.

5) Publications

To be updated…


Tags:
Created by Administrator on 2020/02/03 18:32
    
This wiki is licensed under a Creative Commons 2.0 license
XWiki Enterprise 6.2.2 - Documentation