Land Sub-Project

Last modified by s2s_wiki on 2020/05/21 10:29

Scientific Objectives 

The Land sub-project has proposed the following questions to address:

  1. What is the impact of the observing system on land initialization and S2S forecasts? 
  2. How well are the coupled land/atmosphere processes represented in S2S models? 
  3. How might anomalies in land surface states contribute to extremes?

Activities for 2019-2023 (S2S Phase 2)

What is the impact of the observing system on land initialization and S2S forecasts? 

LS4P (leads: Yongkang Xue [UCLA], Tandong Yao [TPE], Aaron Boone [MétéoFrance]): The project intends to address two questions: (1) What is the impact of the initialization of large scale surface and subsurface land temperature and snow pack, including the aerosol in snow, in climate models on the S2S prediction over different regions? (2) What is the relative role and uncertainties in these land processes versus in SST in S2S prediction and how do they synergistically enhance the S2S predictability? This project focuses more on the process understanding and predictability rather than operational S2S prediction, but nevertheless informs improvements in prediction. Focus is on anomalies over elevated terrain such as the Tibetan Plateau, the Rockies and the Andes, where surface temperature anomalies have a stronger influence on the free troposphere and large-scale circulation. 

GLACE-ESM (lead: Andrea Alessandri [KNMI/ECMWF]): This project focuses on the role of vegetation (modeled interactively) as a source of land surface anomalies, climate variability and predictability within the classical land surface predictability context. S2S is at the short end of the range of timescales to be investigated. The project is currently stalled due to several issues. The inclusion of Earth system processes over land (from CMIP6 developments) turns out to be quite disparate among the different modeling frameworks, even just within Europe, complicating the structuring of sensitivity experiments built around land surface properties such as vegetation. More thoughtful discussion, preliminary experimentation and international coordination are needed. Furthermore, most modeling centers have been dedicating their resources to CMIP6 production. Funding has also been a concern: participating centers are submitting a H2020 proposal and revising the schedule for a late 2020 start.

SNOWGLACE (leads: Yvan Orsolini [NILU] and Jee-Hoon Jeong [CNU]): This is a WCRP-level project to evaluate how individual state-of-the-art dynamical forecast systems vary in their ability to extract forecast skill from snow and sea ice cover initialization, and estimate the impact of snow on surface air temperature and circulation at subseasonal time-scales for boreal winter. Results using the Norwegian Climate Forecast Model have been recently published (doi:10.1029/2019JD030903). 

Within CMIP6 LS3MIP, LFMIP-Pobs will replicate the GLACE2 subseasonal prediction study by extending the number of participating models, period of forecasts, and number of start dates. Initial land states are derived from offline generated pseudo-realistic land states, similar to the LMIP set-up, and are perturbed by shuffling start years (retaining the calendar day) for every member.

Land Data Assimilation is being performed at an increasing number of operational forecast centers to improve initialization of land states for S2S prediction. ECMWF currently assimilates scatterometer and SMOS data operationally and plans to add SMAP to the operational suite in 2020. The Canadian Land Data Assimilation System (CaLDAS) has been operational since 2017 to better represent the land surface initial states in environmental prediction and assimilation systems using an external land surface modeling system and an ensemble Kalman filter (EnKF) methodology. NASA/GMAO also has an operational assimilation system for satellite data. NCEP plans to implement one as part of its multi-year transition to the Unified Forecast System (UFS), which has for the first time been opened to community open-source model and software development.

How well are the coupled land/atmosphere processes represented in S2S models? 

GEWEX has a number of ongoing efforts of relevance to S2S prediction, particularly in the area of model evaluation. Several benchmarking efforts aim to evaluate and improve modeling of the land surface and land-atmosphere interactions. 

  • The Protocol for the Analysis of Land Surface models (PALS) is evolving into, aiming to provide a web-based testbed and workflow environment to aid standardization of land model evaluation, with a particular focus on a priori benchmarking. 
  • PALS has become closely linked in a software sense with the International Land Model Benchmarking (ILAMB) project, an effort with origins in CMIP designed to improve the performance of land models and, in parallel, improve the design of new measurement campaigns to reduce uncertainties associated with key land surface processes. 
  • The PALS Land Surface Model Benchmarking Evaluation Project (PLUMBER and its 2nd phase: PLUMBER2) aim to gauge whether land models are performing as well as they could, given the information they're provided with, and if not, identify practical avenues for model improvement. The first phase of PLUMBER was a site-based model comparison experiment that used out-of-sample empirical models as benchmarks to define model performance expectations. PLUMBER2 is expanding the number of locations from ~20 to ~200, improving the benchmarking of empirical models, and investigating diagnostic approaches to identify model weaknesses. 

PALS, PLUMBER and ILAMB are global standards for model development, with parallel broadening of scopes to include land-atmosphere interactions.

The NOAA Climate Program Office is supporting a Climate Process Team focused on parameterizing the effects of sub-grid land heterogeneity on the atmospheric boundary layer and convection.  Called CLASP (Coupling of Land and Atmospheric Subgrid Parameterizations, lead: Nate Chaney [Duke U.]), the work aims to explicitly link heretofore separate efforts to represent subgrid atmospheric turbulent and convective parameterizations with land surface heterogeneity. Such a novel interaction scheme should improve model performance and prediction of weather and climate elements affects by land-atmosphere interaction. Five US global modeling centers are involved along with researchers in academia. 

Hydro-JULES is a multi-institutional UK effort to merge land surface modeling and distributed hydrologic modeling to aid forecasting of all elements of the water cycle.  

How might anomalies in land surface states contribute to extremes?

Several of the projects listed in A include examination of extremes. LS4P has placed initial focus on the role of temperature anomalies over the Tibetan Plateau. Early results show significant and systematic shifts in the Northern Hemisphere Rossby wavetrain associated with those anomalies, as well as model biases, over Tibet. These have robust and persistent remote responses in many regions in terms of temperature anomalies, but precipitation responses have not been found to be reliably driven by temperature anomalies over Tibet.

There are also a number of singular scientific investigations that are generally well directed and well executed, if not well coordinated. Recently, the 2018 drought and heatwave over northern Europe has been an event of opportunity that has spawned a number of studies and papers into predictability and prediction of mid-latitude extremes with a distinct focus on the role of the land surface. Petch et al. (2020) examined the relative roles of soil moisture and SST in contributing the heatwave over the UK (doi:10.1002/asl.948), finding approximately similar magnitudes for imacts over Britain. Dirmeyer et al. (2020; submitted) examines the heatwave from a process framework using in situ observations and reanalysis to find much of northern Europe entered an unprecedented realm of positive land-atmosphere feedbacks where prolonged dry soils amplified temperatures via increased surface sensible heating, reinforcing the effects of circulation anomalies.

Proposed Activities for 2020 (including timeline and deliverables)

  • There will be a 4-day Land Surface Modelling Summit held during the week of 7-11 September 2020 in Oxford, UK. Its purpose is to bring together the international community of land surface modelers to address shared problems, priorities and ways forward. Two of the four themes are directly related to S2S prediction: information Biophysical Impacts on Weather and Climate; (ii) Water Cycle, Floods, Droughts and Water Resources. The Summit will provide an excellent forum to engage the global modeling community on issues of relevance to the role of the land surface in S2S prediction.
  • USGCRP is establishing a U.S. GEWEX Office, somewhat analogous to the US CLIVAR effort parallel with WCRP/CLIVAR. USGCRP has spent the last year querying the US scientific community to define the scope and goals of US GEWEX, and to ensure a supportive, coordinated role with respect to WCRP/GEWEX. The plan is to kick-start the effort with a focus on land-atmosphere interactions, including their predictability and role in water cycle extremes. There is hope to use this new platform to better coordinate US funding agencies, likely toward an improved distributed measurement effort in the vein of a distributed field campaign that can lead to improved process understanding and coupled land-atmosphere modeling across climatological moisture and temperature gradients that are well-defined over North America. U.S. GEWEX is provisional for the next couple of years, after which it would become a permanent element of USGCRP pending favorable review.
  • Other projects are ongoing as described above.

Linkages with WCRP/WWRP WGs & projects 

WWRP is in the middle of their current 7-year plan, which features a number of Action Areas (AA) focused on water and its prediction. AA7 is the Integrated Water Cycle. The stated goal is to improve understanding, observation, and modeling of atmospheric hydrologic processes with a view toward improved estimation and predictions of precipitation. These are obviously linked over land to surface fluxes, and connect intimately with S2S priorities in this area.


LS4P - 




Created by Administrator on 2019/11/28 16:49
This wiki is licensed under a Creative Commons 2.0 license
XWiki Enterprise 6.2.2 - Documentation