MEETING OF REGIONAL MODELING NETWORK

Nassau, Bahamas –September 21st- 23rd,2009

This meeting was convened to allow the regional modeling network to assess the progress of the PRECIS modeling work, to date, to consider further work on the current data that has been obtained, so far and to agree on future work to be undertaken, by the group, in the forthcoming year. With reference to the latter, apart from the country specific scenario production which is expected, the group agreed on the importance of starting to produce peer review articles which can be published and of providing input into the Fifth AR.

In addition, the group met with the Oxford University Centre for Environment who are in partnership with the CaricomCenter for Climate Change in the Caribbean (CCCCC), for the implementation of one component of the regional CARIBSAVE proposal, for the tourism sector and which addresses the development of destinational risk maps for the region. The purpose of the consultation was to provide the basis for future contributions, by the regional group, to the development of the regional risk maps.

Modeling results to date

PRECIS-CARIBE (Actual data)

 HadCM3 50km

 2x30 Years Baseline Ensembles (Control Experiments) 1960-1990 (S)

 3xA2 30 Years de future scenarios 2070-2100 (S)

 B2 30 Years de future scenarios 2070-2100 (S)

 ECHAM4 50km

 Control 1960-1990 31yrs

 A2 1990-2100 114 yrs

 B2 1990-2100 114 yrs

 ECHAM4 25Km(Western & Eastern Caribbean)

 Control 1960-1990 31yrs

 A2 1990-2100 114 yrs

 B2 1990-2100 114 yrs

 ERA with sulphur cycle 50km

 20 Years Reanalysis ERA40 (1969-1989)

 15 Years Reanalysis ERA15 (1979-1993)

 SRES Control hourly with sulphur cycle 50km

 1960-1969 10 yrs

Convección Medio Ambiente, Palacio de Convenciones Julio 2009

In collaboration with INSMET the CCCCC has completed modeling of the CaribbeanBasin, Central America and the Gulf of Mexico at the 50 km and 25 km resolutions. The modeling was done using the PRECIS in conjunction with the HADCM3 and ECHAM4 GCMs.

Current data –proposed analyses

  1. Determining the added value of doing regional modeling
  • Comparing the RCM and GCM outputs using PRECIS existing data to determine the added value of using the RCM on the domain. Only 50km resolution runs have been analyzed, so far. The variables being considered are precipitation and 2m temperature.

Concerns:

  • Spatial bias needs to be done to determine if the added value differs over the domain.
  • Value of changing domain size and resolution to be determined by similar methodology.

Commitments:

  • INSMET: to determine spatial bias using concept discussed andto apply for domain and resolution analysis.
  • CSGM and CIMH to examine data for characterizing bias/value added and to facilitate online discussion of results.

Timeline:

  • Diagrams to be generated and disseminated by early November
  • INSMET will begin writing the paper and addressing the current concerns.
  • Draft of the paper should be available by mid-November. Online discussions to occur.
  1. Extreme climate
  • Examining extreme weather in the Caribbeancaused byclimate change is, sometimes, more useful for determining the impact on the community. The variables being considered are atmospheric temperature and precipitation. Extreme variables are also being considered; frequency of consecutive wet days and dry spells, maximum 5-day rainfall accumulations, SDI, temperature minimum and maximum, among others.
  • Concerns:
  • Seasonality: cannot be considered in extremes analysis because of varying year length.
  • Lack of temperature extremes
  • Need to evaluate model's ability to simulate current/present day extremes and trends. Availability of data (daily, 30 years) limiting factor.
  • Commitments:
  • CIMH: provision of daily temperature and precipitation data for Barbados and any other Caribbean island available for 30 consecutive years to enable validation (to best of current ability).
  • CSGM: has committed to manipulating the observed data from the various sites for comparison with the model outputs.
  • Timeline:

a)CIMH will provide the observational data to CSGM by mid-Oct

b)CSGM will manipulate the observed data into the format for INSMET by November

c)INSMET will begin writing the paper and addressing the current concerns.

d)A draft of the paper should be available by mid-November

  1. Hurricanes in climate runs
  • This research undertaken, mostly by Israel, (INSMET) is to indicate the frequency and track of hurricanes in regional model data (current and future runs). A methodology for determining 'tropical cyclone like vorticities' (TCLVs) from daily modeled data has been developed.

Concerns:

  • No hurricanes in the Eastern part of the Caribbean irrespective of what is driving the model (Had, ECHAM, reanalysis). Suggests a regional model deficiency.
  • Though unlikely, need to rule out that absence of hurricanes not a consequence of the methodology.
  • Changes in number, track and climatology of the hurricanes all need to be examined

Solutions:

  • Do analysis on a dataset with known hurricanes in eastern Caribbean. Perhaps use on season of MM5 real-time forecasting done by INSMET or CIMH.
  • Analyze hurricane development criteria in model to ascertain cause for missing hurricanes e.g. SST, shear and presence of tropical waves at eastern boundary of model.
  • Determine, monthly, the hurricanes which form to assist in determining the seasonality
  • Increase the domain to the east and the north (long term)

Commitments:

  • INSMET to address concerns above
  • CSGM to assist with seasonal analysis.

New work-proposed

  1. Participation in the Coordinated Regional climate Downscaling Experiment
  2. (CORDEX)Project?

Illustration 1: CORDEX domain

The consensus is that we should participate in the modeling aspect so that future modeling activities will yieldcomparabledomain, spatial and temporal resolutions. It was, however, decided that the group would suggest that the domain should be further north (~40N) to account for the subtropical high, an important driver in the Caribbean weatherand that this should be communicated to the CORDEX organizers.

The direction of the modeling activities in the Caribbean should be oriented not only towards academics and research questions but should also provide support for various sectors. There is also a need for inter-model comparisons to better represent uncertainties. This leads to the agreement that the way forward, for the modeling group, is to perform a number of climate runs with various RCMS. These models include the PRECIS model, RegCM, CCSM, WRF and PCM.

The new version of the PRECIS model can facilitate inputs from 12 GCMS and it was previously agreed that 6 of these will be used (2 high, low and middle).

The distribution of the modeling tasks is shown below in the table:

Model / Group / Forcing
RegCM / INSMET
PRECIS / INSMET, CSGM (UWI CH?) / 6 GCM outputs
WRF / CIMH, CSGM / CCSM
PCM / UWI CH
  1. Old/New Emission Scenarios?

The concern with the new emission scenarios is the timeline by which these outputs should be available. In order to support the IPCC AR5 report, results would have to be made available by the end of 2010. Given the constraints of the computational resources currently available and the availability of new emission scenarios, it was decided that the runs should commence with the old A1B scenario.

Ideally it would be best if the climate runs out to the year 2100 were continuous as opposed to time sliced. However, for those institutions opting for the time sliced approach, the CORDEX defined ones will be utilized. These three time slices are decadal hind casts/forecasts for 1980 – 2010, 1990 – 2000, and 2005 – 2035.

All climate runs should complete:

a)Historical runs for the period Jan 1950 – Dec 2005

b)RCP4.5 and RCP8.5 scenario runs for the period Jan 2006 – Dec 2100

  1. Use of WRF model in Operational Mode

INSMET and CIMH have the capacity to run the WRF regional model in an operational mode for the production of short term forecasts. In the case of INSMET outputs from this model inform their disaster management programme of activities, especially during the hurricane season. Their outputs cover the entire region and can be made available, on a real time basis, to meteorological officers and disaster managers, throughout the Caribbean. This could be a powerful tool for the disaster management community, in the Caribbean, in terms of its potential to provide real time accurate forecasts, especially during the hurricane season.

In the case of CIMH, WRF and MM5 have been running in operational mode for approximately 2 years. Outputs from these twice daily runs are provided to the public on our web page at The domains cover the entire Caribbean region and a portion of the North Atlantic. A number of the regional meteorological services use theseoutputs to support their forecasting. Additionally CIMH and CDERA (now CDEMA) have established a cooperative arrangement in which CIMH embeds a meteorologist in CDERA during periods of severe weather during the hurricane season.

The goal of the embedding is to improve CDEMA's scenario analysis. Outputs from the CIMH models, as well as outputs international modeling centers, are utilized by CIMH to provide CDEMA with a set of likely weather scenarios. This system was in placesince last year and proved very useful for Hurricane Dean and a number of systems. This yearthis mechanism was used for Hurricane Bill that passed through the north eastern Caribbean.

In addition, CIMH provides CDEMA with storm surge forecasts based on outputs fromthe TAOS RTFS product developed by Kathy-Ann Caesar and made available to the region throughCCRIF. CIMH, CDEMA and CCRIF have engaged in a programme to educate the region onthe range of products available for the region to support disaster management.

It was agreed that a workshop, for regional meteorological officers, especially those from countries without adequate forecasting capacity and their respective Disaster managers, would be appropriate. This would acquaint them with the availability of the WRF outputs, access through the website, extraction and interpretation of the data derived from the website and its eventual dissemination and use in the Disaster community. It was agreed that if resources were available a three day regional workshop wouldbe held,in November, to provide training to regional meteorological and disaster management personnel in accessing and utilizing this facility.

Inputs into the CARIBSAVE Project

Caribbean Community Climate Change Centre (CCCCC), in collaboration with the Oxford University Centre for Environment (OUCE), have been implementing a pilot phase of one component of a more holistic proposal, CARIBSAVE, which addresses the vulnerability of the regional tourism sector to climate change and seeks to develop adaptation/mitigation responses to build a more climate resilient tourism sector in the region. The component being addressed, on a pilot scale in Negril in Jamaica and Eleuthra, in the Bahamas seeks to develop a Destinational Climate Change Risk Map that will form the basis of integrating climate risks into the sectoral (tourism in this instance) planning process. Development of these risk maps depended on the outputs from the regional PRECIS runs and these were supplied by the regional group to the OUCE collaborators who used the outputs to derive destination specific future climate scenarios for Negril and Eleuthra and, further, transformed these to provide risk profiles for the selected destination. In the discussion between the OUCE and regional network it was emphasized that one of the major considerations in the proposed partnership would be capacity building and to the extent that this was possible it should be given the highest priority. Given the limited scope of the pilot activity and the abbreviated timeline for its completion there wasn’t much opportunity for capacity building and transfer of know how between OUCE and the regional group. However the methodology used by the OUCE for the derivation of the risk maps will be recorded and made available to the CCCCC. It was noted that, in the development of the risk profile of the Eleuthra coastal resources,there was full participation of an officer from the local Met office and that a training manual with a full description of the methodology was now available. In future, to the extent that the regional network could perform these tasks (resources and expertise being available) their services will be fully utilized. The latter will be taken on board when resources for the implementation of the full component become available to develop the risk maps for all Caribbean countries. It was agreed that the partnership with OUCE could be mutually beneficial and some consideration might be given to links that would facilitate post graduate training,at Oxford, for regional scholars in the climate change field.

Oxford undertook to look into the possibility of acquiring a new regional model (CLE???), developed by the EU, so that it could be utilized in the future work of the Caribbean group.

Resources

The regional modeling effort, to date, has been successful but has been implemented with minimal resources. There is now emerging a growing demand for the outputs from these exercises to feed into national/regional activities e.g. SNCs and this will grow in the future. Accordingly, the modeling group to meet these demands will have to be properly resourced to ensure that the necessary capacity to respond is in place (equipment & human resources). It was agreed that the group will put together comprehensive proposal which addresses this capacity need and submit this to the Centre which will then seek to target resources for its implementation. Possible sources of support were DFID (through the CARIBSAVE project) EU, IDB and others.

In attendance for PRECIS discussions

INSMET

  • Dr Abel Centella
  • Mr.Arnoldo Benzanilla
  • Mr.Israel Borrajero

UWI - CSGM

  • Dr Michael Taylor
  • Dr Tannecia Stephenson
  • Mr. Trevor Hall

UWI – Cave Hill

  • Dr John Charlery

CIMH

  • Ms Kim Whitehall

Joining the group for the discussion on CARIBSAVE are:

OUCE

  • Dr Murray Simpson
  • Dr Mark New
  • Dr Carol McSweeney ( Joining the Hadley Centre in October)

Meteorological Office of theBahamas

  • Ms Suzane Russel
  • Mr. Orson Nixon

DFID (Barbados office)

  • Ms Simone Bannister

IDB (Washington office)

  • Mr. Gerard Alleng

Ulric O’D Trotz

1