WCRP-WWRP-THORPEX Model Evaluation and Development Survey - Community Issues


Q5: What would best accelerate progress? Suggestions for new intiatives (new process studies, field campaigns, or new collaborative approaches, eg. international working groups, Climate Process Teams).

It is difficult for small R&D groups that are away from major modelling centers to contribute to progress unless code is modularized and open source.

Need to address how systematic errors come about (resolution, understanding - parameterization).

Lack of marine observations to understand air-sea interactions - surface fluxes (especially ppt and latent heat), SSS.

The nearly un-bending IPCC process that is on a mission to communicate that humans are causing climate change.

Difficulty in accessing international observational datasets.

Lack of suitable observations for model evaluation, especially on long timescales.

Combining modelling community to ice sheet dynamics process study groups is currently ad-hoc (using graduate students).

Disconnect between diagnostics community and modelling community. This leads to diagnostitians to try to develop statistical methods to remove model biases and errors and model development communities making their improvements independently.

Enhance the connection between different modeling communities (climate, NWP) and between modeling, process studies and observations.

Huge gap between the process studies and modelling communities and insufficient support of field experiments from the modelling community.

Improve links between met services and geosciences university departments to exploit ground-based GPS observations and for land process research and observations.

Better integrate the NWP and climate community and global and regional model development.

The US should heavily invest in at least one combined NWP/climate center (eg combined functions of NCEP and GFDL) in which global and regional modeling are fully integrated and in which there is both a critical mass of in-house developers on long-term contracts and adequate funding for entraining partners from the outside research community into model development projects. Currently there is over-investment in field experiments and new satellite instruments compared to model development that, combined with inadequate model integration, has left modeling efforts behind ECMWF and the UKMO/Hadley Centre.

Making key observational data available with evaluation of quality and reliability.

More international access to observational climate and weather data generated in Europe.

Bridging the gap between LES/CRM and GCMs, eg as led by GCSS. Few LES/CRM studies go beyond stating their superiority to GCMs and these tools are not exploited to explore possible improved parameterization approaches.

Observations: more needed to retrieve cloud scale vertical velocities from radar, to constrain the 3D water vapor field at high vertical resolution, to develop more accurate retrievals of ice water content in clouds.

Resourses to better exploit: CloudSat/CALIPSO, VOCALS, TRMM latent heating retrievals.

Many people funded to do cloud research, only a handful develop parameterizations that are used in operational models.

The IPCC cycle should be longer - current ~6yr cycle leaves 2yrs for model runs, 2yrs for evaluation and only 2yrs for model development.

Existing efforts to define metrics for climate models are biased towards well observed fields (eg dynamical fields), not those important for climate (cloud structure, water vapor structure, aerosols, sea ice), and based on simple spatial distributions of monthly fields that are not useful for extremes, climate sensitivity, regional changes, sea ice decline, etc. Unlike NWP skill scores, these do not identify which models are best for predicting climate.

The gap between those working in statistical predictability studies, observationalists that provide the initial/forcing conditions and those doing parameterization development - need an integrated project framework.

WG on non-hydrostatic modelling.

Disconnect between those working on the consequences of different convective schemes in idealized models and those looking at GCM sensitivity.

Need to develop techniques for observing deep ocean temperature and stability distributions to complement Argo for the upper ocean. Also need improved analysis of upper and deep ocean observations to minimise the impacts of instrumental biases and gaps (eg sophisticated reanalysis techniques - adjoint systems).

Continued support for the PCMDI model archive and the IPCC process for their scientific benefits as well as their societal impacts.

The modelling community has a trend to explore untreaded territories (eg decada prediction) at the expense of working on persisting unsolved problems (eg ENSO prediction, tropical biases). The community needs a balance between these two aspects and encourage and supprot diligent, unglamorous efforts at tackling known tough problems that may take a long time to be solved.

Insufficient coordination between different ecosystem-related scientific communities (plant physiologists, experimental field ecologists, micrometeorologists, hydrologists, etc) means that the information base is inadequate for resolving extraordinarily contetious issues such as the ecosystem level effects of CO2 concentration on both carbon and water cycling. In particular, we are reliant on a very small handful of forest Free-Air Carbon Dioxide Enrichment (FACE) experiments that are confined to temerate forests (and generally confined to the USA) that strongly focus on the carbon cycle.The effort needs to be scaled up with strategic experiments located in different biomes and with a more serious involvement of modellers in the design and interpretation of the experiments.

The current configuration of the international global change programmes is unfit for purpose, in particular the separation between WCRP and IGBP. There is a need for a unified body with the ear of national funding agencies.