Royal Dutch Meteorological Institute; Ministery Of Infrastructure And The Environment

 

 

PROCEED project

PROCess-based sEamless development of useful Earth system predictions over lanD (PROCEED)


 

The lack of observations to constrain the model complexity over land has determined, so far, the development of different prediction systems for different time scales. While benefit from daily verification, the models that are developed for short time-scales (Weather to seasonal climate predictions) include only that part of the surface and near-surface variability for which observations are available and that can be suitably modelled/initialized in order to positively contribute to the forecasts (verification-based approach). Therefore, to limit prediction errors, short time-scale models do not include those processes related to vegetation and their seasonal, interannual and sub-grid variability.There is no scientific basis to draw artificial boundaries between synoptic-scale numerical Weather forecast, seasonal prediction, ENSO prediction, decadal prediction and climate change. The Earth System exhibits a wide range of dynamical, physical, biological, and chemical interactions involving spatial and temporal variability continuously spanning all weather/climate scales. Therefore, the gaps between the models used for short-term prediction and the latest developments in modelling the land surface are believed to considerably limit the current level of performance and usefulness of predictions.

 

 

Objectives

  1. Develop novel observational constraints to land-climate interactions and feedbacks: Develop a comprehensive array of up-to-date state-of-the-art land data by exploiting latest developments in land monitoring services. The emerging land data sets will allow derivation of a novel set of observational constraints on land-climate interactions and biophysical feedbacks.
  2. Understand the land limitations that are affecting current prediction models: Identify the specific unresolved land surface-vegetation processes that affect the predictions from the short subseasonal-to-seasonal time-scales and that need to be adequately represented to overcome the limitations of the system across scales.
  3. Enhance Earth System predictions across scales by obtaining a practicable seamless development of verifiable land surface processes: Introduce in a seamless way those land vegetation processes that improve the performance and usefulness of the predictions by overcoming the artificial boundaries between short term and longer term climate predictions over land. By one hand, we want to extend the selected processes from the longer to the shorter time scales. In return, the verification at the short time-scales will provide knowledge back to the longer time-scales to better constrain the land processes (See Picture below).
  4. Exploit performance/usefulness of improved Earth System predictions over land. Provide valuable information to end-users thanks to the improved predictions over land. By actively collaborating with power grid operators in Europe contribute in developing robust climate services for the energy sector.

The two-way seamless modeling strategy across time-scales developed in PROCEED.

 

MARIE SKLODOWSKA CURIE ACTIONS European Contract H2020-MSCA-IF-2015 704585

Granted Researcher: Andrea Alessandri (KNMI) Institution: KNMI – Royal Netherlands Meteorological Institute

Partner Organization: ECMWF – European Centre for Medium-range Weather Forecasts

Two Years and half Project – Start date 1 January 2017    

Advisory committe: Bart van den Hurk (KNMI), Gianpaolo Balsamo (ECMWF), Franco Molteni (ECMWF)

 

 

Work Plan & Packages

The overall strategy of the workplan is outlined in following Figure, showing the graphical presentation of the work packages with their main interdependencies. The different activities described in the WPs are closely connected and the information will flow from one WP to another as reported in the figure.


A fundamental contribution within PROCEED will be brought by the new climate and land surface observed data that will be collected and analysed in WP1. WP1 will provide the knowledge to go beyond state-of-the-art in WP2, where a verification-based analysis will be applied to evaluate the performance and current limitations of the latest version of the ECMWF forecasting system13. The gain in potential predictability across scales that is obtained representing new land vegetation processes in an extended version of the ECMWF seasonal prediction system, developed for climate variability/change research22, will be assessed in WP3. The synergy between verification-based and process-based analyses will be exploited in WP4 where the seamless development of the prediction system will be performed. Finally, in WP5 the evaluation of forecast improvements will be performed across scales with particular focus to the usefulness for the energy sector. The overall strategy of the work plan is outlined in Figure 1, showing the graphical presentation of the work packages with their main interdependencies. The different activities described in the WPs are closely connected and the information will flow from one WP to another as reported in the figure. The activities of the 5 work packages is detailed below:

WP1) Develop novel observational constraints to land-climate interactions and feedbacks.
Milestone 1.1: Delivery of datasets for analysis, model initialization and evaluation that will also serve WP2, WP3, WP4 and WP5.
Deliverable 1.1: Report/scientific paper on the novel observational constraints to land-climate interactions and feedbacks.

WP2) Verification-Based analysis: performance and limitations of ECMWF prediction system over land.
Deliverable 2.1: Report/scientific paper on the performance and limitations of State-of-the-Art forecasts over land from latest ECMWF prediction system.

WP3) Process-Based analysis: land vegetation processes contribution to potential predictability in the ECMWF system.
Milestone 3.1: Delivery of a set of potential predictability experiments with newly introduced land vegetation processes.
Deliverable 3.1: Report/scientific paper(s) on the evaluation of the potential contribution of land vegetation processes to potential predictability.

WP4) Process based seamless development of climate predictions over land.
Milestone 4.1: Delivery of a new version of the prediction system with a seamless development of the land surface-vegetation processes.

WP5) Improved Earth System predictions across scales: skill and value for the energy sector.
Milestone 5.1: Phase 1 of the improved Earth System predictions across scales completed.
Milestone 5.2: Delivery of the full set (Phase 2) of improved Earth System predictions across scales.
Deliverable 5.1: Report/scientific paper(s) on the evaluation of improved Earth System predictions across scales: forecasts skill and value for energy predictions.

 

 

Work performed and achievements

 

WP1) Develop novel observational constraints to land-climate interactions and feedbacks

A new generation satellite data products [Essential Climate Variables (ECVs) over land] released as part of the ongoing implementation of the Global Climate Observing System (GCOS) have been collected. The data (Leaf Area Index [LAI], albedo, Land Cover, Green Vegetation cover, soil moisture) are suitable to be used to validate and better constrain the modeling of land-surface processes.
The observational ECV data over land have been analyzed to understand processes driving vegetation and surface albedo variability and the coupling/feedbacks with atmosphere. This analysis is producing a paper publication under submission in Nature Geoscience. Furthermore, the analysis has led to contribution to a paper, currently in revision in JGR biogeosciences, aimed at the identification and characterization of onset/offset of the vegetation growing season.

Milestone 1.1: Delivery of datasets for analysis, model initialization and evaluation that will also serve WP2, WP3, WP4 and WP5

-> “Delivery of datasets for analysis, model initialization and evaluation that will also serve WP2, WP3, WP4 and WP5” accomplished

Follow this link for details of the Collected Observational datasets in Milestone 1.1 and accessibility

 

Deliverable 1.1 Report/scientific paper on the novel observational constraints to land-climate interactions and feedbacks

-> Scientific paper “Different Signatures of Land-Albedo Feedback on the Northern Hemisphere Surface Warming” by Alessandri and co-authors, Under Submission Nature Geosciences.
Full paper will be shared in this page at final acceptance of the paper. Please contact me if you want to have a look at draft manuscript.

-> Contribution to the observational paper on the development of new tools for the identification and characterization of onset/offset of the vegetation growing season: D. Peano, S. Materia, A. Collalti, A. Alessandri, A. Anav, A. Bombelli, S. Gualdi, 2019: Variability of simulated and observed growing season onset/offset. Under revision in JGR biogeosciences.
Full paper will be shared at this link at final acceptance.


 

WP2) Verification-Based analysis: performance and limitations of ECMWF prediction system over land.

The performance and limitations of the latest version of the operational ECMWF seasonal prediction system (System5, released at the end 2018; Johnson et al., 2019; Stockdale et al., 2019) has been evaluated and compared with previous version (System 4; Molteni et al., 2011) by exploiting the observational data from WP1. Thanks to this analysis it has been possible to identify major gaps in the representation of land surface/vegetation processes in the ECMWF system. The knowledge about the limitation in the representation of land-vegetation in the ECMWF system will be fundamental knowledge for the modeling developments in WP3-WP4.

Deliverable 2.1 Report/scientific paper on the performance and limitations of State-of-the-Art forecasts over land from latest ECMWF prediction system

-> Report on the “Evaluation of ECMWF seasonal climate prediction performance over Land: System 5 vs. System4”

Outcomes of the analysis described in this report will be published in peer-reviewed paper for the scientific community that are currently in preparation:
- A. Alessandri, and co-authors: Evaluation of ECMWF seasonal climate prediction performance over Land: System 5 vs. System4. In Preparation.

 


WP3) Process-Based analysis: land vegetation processes contribution to potential predictability in the ECMWF system.

A set of retrospective forecasts and potential predictability experiments have been performed with and without the representation of the new land vegetation processes. The sensitivity to the following key processes has been accomplished:
-Realistic vegetation/land cover: the HTESSEL formulation prescribing constant vegetation coverages is replaced by a modified model version that allows vegetation/land cover fractional coverage to change as a function of Leaf Area Index
-Interactive surface albedo: the current parameterization, prescribing time-invariant blended albedo for each grid point, has been replaced with an interactive albedo scheme discriminating between vegetation and soil. A parameterization of the bare soil albedo is then deduced as a function of soil water content.

 

Milestone 3.1 Delivery of a set of potential predictability experiments with newly introduced land vegetation processes

-> Set of potential predictability experiments with newly introduced land vegetation processes accomplished.

 

Deliverable 3.1 Report/Scientific paper(s) on the evaluation of the potential contribution of land vegetation processes to potential predictability

        -Realistic vegetation/land cover:
-> Paper on Multi-scale enhancement of climate prediction over land by increasing the model sensitivity to vegetation variability in EC-Earth:
A. Alessandri, F. Catalano, M. De Felice, B. Van Den Hurk, F. Doblas Reyes, S. Boussetta, G. Balsamo, and P. Miller, 2017: Multi-scale enhancement of climate prediction over land by increasing the model sensitivity to vegetation variability in EC-Earth, Clim. Dyn, 49(4), 1215-1237. doi:10.1007/s00382-016-3372-4, link to-readcube document http://rdcu.be/kJuG


        -Interactive surface albedo:
-> Report on the “Development of new interactive surface albedo scheme in EC-Earth ”.

Outcomes of the analysis described in this report will be part of peer-reviewed publications for the scientific community that are currently in preparation:
- A. Alessandri, and co-authors, 2019: Development of a new process-based albedo parameterization in EC-Earth. In Preparation.
- F. Catalano, Alessandri A., and co-authors Process-based enhancement of climate predictions in EC-Earth due to improved soil-moisture albedo feedback. In preparation.

 

 

WP4) Process based seamless development of climate predictions over land.

The knowledge obtained from the verification-based (WP2) and process-based (WP3) approaches has driven the seamless development in this WP of an improved land modeling in EC-Earth and ECMWF System5. The improved representation of vegetation/land cover and of the interactive surface albedo has been included in the latest EC-Earth release (version 3.3.2) and that will be used for CMIP6 simulations (http://www.ec-earth.org/cmip6/ec-earth-in-cmip6/). This task has been accomplished in the framework of the Land and Vegetation working group of EC-Earth.
On the other hand, the improved representation of land surface processes that has been identified to potentially improve prediction skill has been included in the latest ECMWF seasonal prediction system (SEAS5). This task has been accomplished via ECMWF account and as collaborative effort with colleagues at ECMWF [relevant people from the Earth System predictability department and of the Research and Development department (M. Balmaseda, T. Stockdale, G. Balsamo, G. Arduini, S. Boussetta)].

 

Milestone 4.1 Delivery of a new version of the prediction system with a seamless development of the land surface-vegetation processes

-> Delivery of new version of EC-Earth accomplished: contribution to EC-Earth release 3.3.2 with improved representation of land surface/vegetation/land cover and of the interactive surface albedo (http://www.ec-earth.org/cmip6/ec-earth-in-cmip6/).

-> Delivery of new version of ECMWF prediction system accomplished: delivery of branch:daaa_SB43R1_daaa_CY43R1_SYS5lowresStephanie with improved representation of land-surface processes in the latest cycle (CY43R1) of the SEAS5 forecasting system at TCO199 resolution (https://confluence.ecmwf.int/display/~daaa/Release+of+PROCEED+SEAS5+branch)

 


WP5) Improved Earth System predictions across scales: skill and value for the energy sector.

An evaluation of the Earth system prediction improvements obtained from the developments in WP4 is accomplished globally across different time scales with a particular focus is on the predictability of surface climate over European domain.
The effects of the improved representation of land surface-vegetation that has been included in the latest ECMWF seasonal prediction system have been evaluated.
In a set of historical simulations performed following the CMIP6 protocol, the effects of the new interactive soil albedo scheme in the latest EC-Earth release from WP4 have been assessed.
Creating and exploiting an international network of collaborations with research and end-users, it has been performed an evaluation of the potential usefulness of the improved climate predictions. A comprehensive assessment of the skill, probabilistic quality and potential economic value for end users in the energy sector has been accomplished, leading to collaborative efforts that have been published in peer-reviewed journal Climate Dynamics. Furthermore, in another collaborative peer-review paper it has been proposed an approach to evaluate how the latest developments in Seasonal climate forecasts can provide a useful prediction for the average photovoltaic (PV) power production over European regions. It is of great importance to understand and predict Indian Summer Monsoon Rainfall seasonal anomalies because of their teleconnected impacts over Euro-Mediterranean domain. To this aim, contribution to the application of an advanced causal discovery algorithm has been accomplished.

Milestone 5.1 Phase 1 of the improved Earth System predictions across scales completed

-> Preliminary set of simulations accomplished with EC-Earth and ECMWF (SEAS5) to have analysed a small subset before engaging full set of simulations.

 

Milestone 5.2 Delivery of the full set (Phase 2) of improved Earth System predictions across scales

-> New set of Seasonal climate predictions (1982-2014) with improved version of SEAS5 (daaa_SB43R1_daaa_CY43R1_SYS5lowresStephanie) accomplished.

-> Historical simulations (1900-2014) with new version of EC-Earth accomplished

 

Deliverable 5.1 Report/scientific paper(s) on the evaluation of improved Earth System predictions across scales: forecasts skill and value for energy predictions

-> Paper on the evaluation of latest developments in Seasonal climate prediction in terms of probabilistic performance and potential economic value for energy load forecasting over Italy (collaboration with Italian TSO – TERNA):
A. Alessandri, M. De Felice, F. Catalano, J-Y. Lee, B. Wang, D-Y. Lee, J-H. Yoo, A. Weisenheimer, 2018: Grand European and Asian-Pacific multi-model seasonal forecasts: maximization of skill and of potential economical value to end-users, Clim. Dyn., 50: 2719. https://10.1007/s00382-017-3766-y (SHERPA/RoMEO Green policy of Clim Dyn); Link to open access repository at the University of Oxford: https://ora.ox.ac.uk/objects/uuid:b469a38f-a347-431f-99ac-862e93c8c3c3

-> Contribution to paper on the scoping of the usefulness of latest developments in Seasonal climate forecasts for the prediction of renewable energy (solar power generation) over Europe: M. De Felice, M. B. Soares, A. Alessandri, A. Troccoli, 2019: Scoping the potential usefulness of seasonal climate forecasts for solar power management, Accepted, https://10.1016/j.renene.2019.03.134. link at open access preprint: https://eartharxiv.org/vfn35

-> Report on the “Enhancement of seasonal climate prediction in ECMWF SEAS5 by representing realistic vegetation-cover variability”.

Outcomes of this analysis will be published in a peer-reviewed paper for the scientific community that is currently in preparation:
- A. Alessandri, and co-authors: Enhancement of seasonal climate prediction in ECMWF SEAS5 by representing realistic vegetation-cover variability. In Preparation.

-> Report on the “Effects of improved land-vegetation representation in EC-Earth: sensitivity to the new soil albedo scheme with dependence on soil moisture”.

Outcomes of the analysis described in this report will be part of peer-reviewed publications for the scientific community that are currently in preparation:
- A. Alessandri, and co-authors, 2019: Development of a new process-based albedo parameterization in EC-Earth. In Preparation.
- R. Doescher, and co-authors: EC-Earth in CMIP6. Under submission in Geosci. Model Dev.

-> Contribution to paper on the development of innovative techniques for the prediction of the Indian Summer Monsoon and relationship with Euro-Mediterranean climate in boreal summer: G. Di Capua, M. Kretschmer, J Runge, A. Alessandri, R. Donner, B. van den Hurk, R. Vellore, R. Krishnan, and D. Coumou, 2018: Long-lead empirical forecasts of the Indian Summer Monsoon Rainfall based on causal precursors. Accepted in Weather and Forecasting. https://doi.org/10.1175/WAF-D-19-0002.


*****

All simulations and related output data in PROCEED were produced using the ECMWF's High Performance Computing Facility (HPCF) in Reading (UK). Therefore the immediate access to all results and data generated by the project is guaranteed by accessing the dedicated computing facilities at ECMWF via shared directory and/or mars archiving system.

Post-processed data for specific collections of variables and over specific domains will be provided under request in order to facilitate their use and reduce the difficulties of transferring and storing huge amount of data.

The scripts for the analysis and verification will be provided under request to all interested users.

*****

 

This project has received funding from the European Union’s Horizon 2020 research and innovation programme under grant agreement No 704585.