• Campbell, J. D., , M. A. Taylor, , T. S. Stephenson, , R. A. Watson & , and F. S. Whyte, 2011: Future climate of the Caribbean from a regional climate model. Int. J. Climatol., 31, 18661878, doi:10.1002/joc.2200.

    • Search Google Scholar
    • Export Citation
  • Cashman, A., , L. Nurse & , and J. Charlery, 2010: Climate change in the Caribbean: The water management implications. J. Environ. Dev., 19, 4267.

    • Search Google Scholar
    • Export Citation
  • Centella, A., , A. Bezanilla & , and K. Leslie, 2008: A study of the uncertainty in future Caribbean climate using the PRECIS regional climate model. Community Caribbean Climate Change Centre Tech. Rep., 16 pp.

    • Search Google Scholar
    • Export Citation
  • Charlery, J. & , and L. Nurse, 2010: Areal downscaling of global climate models: An approach that avoids data remodeling. Climate Res., 43, 241249, doi:10.3354/ cr00875.

    • Search Google Scholar
    • Export Citation
  • Cox, P. M., , R. A. Betts, , M. Collins, , P. P. Harris, , C. Huntingford & , and C. D. Jones, 2004: Amazonian forest dieback under climate-carbon cycle projections for the 21st century. Theor. Appl. Climatol., 78, 137156.

    • Search Google Scholar
    • Export Citation
  • Curtis, S. & , and D. W. Gamble, 2008: Regional variations of the Caribbean mid-summer drought. Theor. Appl. Climatol., 94, 2534, doi:10.1007/s00704-007-0342-0.

    • Search Google Scholar
    • Export Citation
  • ECLAC, cited 2012: Review of the economics of climate change in the Carribean. [Available online at www.eclac.cl/cgi-bin/getprod.asp?xml=/portofspain/noticias/paginas/4/34444/P34444.xml&xsl=/portofspain/tpl-i/p18f.xsl&base=/portofspain/tpl-i/top-bottom.xsl.]

    • Search Google Scholar
    • Export Citation
  • Houghton, J. T., , Y. Ding, , D. J. Griggs, , M. Noguer, , P. J. van der Linden, , X. Dai, , K. Maskell & , and C. A. Johnson, Eds., 2001: Climate Change 2001: The Scientific Basis. Cambridge University Press, 881 pp.

    • Search Google Scholar
    • Export Citation
  • Jones, R. G., , M. Noguer, , D. Hassell, , D. Hudson, , S. Wilson, , G. Jenkins & , and J. Mitchell, 2003: Workbook on generating high resolution climate change scenarios using PRECIS. UNDP, GEF, and Met Office Hadley Centre Manual, 32 pp.

    • Search Google Scholar
    • Export Citation
  • Karmalkar, A. V., , M. A. Taylor, , J. Campbel l, , T. Stephenson, , M. New, , A. Centella, , A. Bezanilla & , and J. Charlery, 2013: A review of observed and projected changes in climate for the islands in the Caribbean. Atmósfera, 26, 283309.

    • Search Google Scholar
    • Export Citation
  • Magaña, V., , J. A. Amador & , and S. Medina, 1999: The midsummer drought over Mexico and Central America. J. Climate, 12, 15771588.

  • McClean, R. F. & , and A. Tysban, 2001: Coastal zones and marine ecosystems. Climate Change 2001: Impacts, Adaptation, and Vulnerability, J. J. McCarthy et al., Eds., Cambridge University Press, 343379.

    • Search Google Scholar
    • Export Citation
  • Mearns, L. O., , F. Giogi, , P. Whetton, , D. Pabon, , M. Hulme & , and M. Lal, 2003: Guidelines for the use of climate scenarios developed from regional climate model experiments. IPCC and TGCIA, 38 pp. [Available online at www.ipcc-data.org/guidelines/dgm_no1_v1_10-2003.pdf.]

    • Search Google Scholar
    • Export Citation
  • Mitchell, T. D., , T. R. Carter, , P. D. Jones, , M. Hulme & , and M. New, 2004: A comprehensive set of high-resolution grids of monthly climate for Europe and the globe: The observed record (1901–2000) and 16 scenarios (2001–2100). Tyndall Centre for Climate Change Research Working Paper 55, 30 pp.

    • Search Google Scholar
    • Export Citation
  • Nakićenović, N. & , and R. Swart, Eds., 2000: Special Report on Emissions Scenarios. Cambridge University Press, 599 pp.

  • Nurse, L., 2011: The implications of global climate change for fisheries management in the Caribbean. Climate Dev., 3, 228241.

  • Singh, B., 1997: Climate changes in the greater and southern Caribbean. Int. J. Climatol., 17, 10931114.

  • Solomon, S., , D. Qin, , M. Manning, , M. Marquis, , K. Averyt, , M. M. B. Tignor, , H. L. Miller Jr. & , and Z. Chen, Eds., 2007: Climate Change 2007: The Physical Science Basis. Cambridge University Press, 996 pp.

    • Search Google Scholar
    • Export Citation
  • Taylor, M. A., and Coauthors, 2007: Glimpses of the future: A briefing from the PRECIS Caribbean Climate Change Project. Caribbean Community Climate Change Centre, 24 pp.

    • Search Google Scholar
    • Export Citation
  • Taylor, M. A., , T. S. Stephenson, , A. Owino, , A. A. Chen & , and J. D. Campbell, 2011: Tropical gradient influences on Caribbean rainfall. J. Geophys. Res., 116, D00Q08, doi:10.1029/2010JD015580.

    • Search Google Scholar
    • Export Citation
  • Taylor, M. A., , F. S. Whyte, , T. S. Stephenson & , and J. D. Campbell, 2013: Why dry? Investigating the future evolution of the Caribbean low level jet to explain projected Caribbean drying. Int. J. Climatol., 33, 784792, doi:10.1002/joc.3461.

    • Search Google Scholar
    • Export Citation
  • Wigley, T. M. L. & , and B. D. Santer, 1993: Future climate of the Gulf/Caribbean basin from atmospheric general circulation models. Climatic Change in the Intra-Americas Sea, G. A. Maul, Ed., Chapman and Hall, 3154.

    • Search Google Scholar
    • Export Citation
  • View in gallery

    (a) Caribbean domains—Big domain at 50 km (all Caribbean, Central America, southern United States, and northern South America) and two smaller domains at 25 km: western Caribbean (red) and eastern Caribbean (green). Also shown are the land–sea mask of the (b) HadAM3H GCM and (c) PRECIS RCM for the big domain, and (d) topography for an 18°N cross section through the big domain as represented by Global 30-arc-second elevation dataset (GTOPO 30) quasi observations (green), HadAM3H (red), and PRECIS (black).

  • View in gallery

    Mean seasonal differences [model – Climatic Research Unit (CRU )] in (a),(b) temperature (°C) and (c),(d) precipitation (%) with respect to the (CRU ) database (Mitchell et al. 2004) for the period of May–Oct 1961–89. Data for (a),(c) HadAM3P and (b),(d) PRECIS are interpolated to the CRU 0.5° grid.

  • View in gallery

    Mean changes in the annual mean 1.5-m temperature for 2071–99 with respect to 1961–89 as simulated by (left) HadAM3P and (right) PRECIS for the SRES A2 scenario.

  • View in gallery

    Annual mean changes in precipitation (%) for 2071–99 with respect to 1961–89 as simulated by (left) HadAM3P and (right) PRECIS. Top (bottom) shows results for the B2 (A2) emission scenario.

  • View in gallery

    Number of simulations projecting precipitation increase for the 2080s (left) annually, and during the (middle) wet season and (right) dry season.

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 79 79 5
PDF Downloads 60 60 2

The Precis Caribbean Story: Lessons and Legacies

View More View Less
  • 1 Climate Studies Group Mona, Department of Physics, University of the West Indies, Mona, Jamaica
  • | 2 Instituto de Meteorologia de la Republica de Cuba, Habana, Cuba
  • | 3 Department of Computer Science, Mathematics and Physics, University of the West Indies, Cave Hill, Barbados
  • | 4 Instituto de Meteorologia de la Republica de Cuba, Habana, Cuba
  • | 5 Climate Studies Group Mona, Department of Physics, University of the West Indies, Mona, Jamaica
  • | 6 Instituto de Meteorologia de la Republica de Cuba, Habana, Cuba
  • | 7 Climate Studies Group Mona, Department of Physics, University of the West Indies, Mona, Jamaica
  • | 8 Department of Infrastructure, Anton de Kom University of Suriname, Paramaribo, Suriname
© Get Permissions
Full access

By the beginning of the current century, there was heightened recognition that the Caribbean is highly vulnerable to the effects of climate change. Yet, there was very little climate change science information for the region and at the scale of the small islands that make up most of the region. To fill the gap, a group of regional scientists representing three institutions and four territories (Barbados, Belize, Cuba, and Jamaica) initiated a project to provide dynamically downscaled climate change information for the Caribbean. The Providing Regional Climates for Impacts Studies (PRECIS)-Caribbean initiative was premised on a shared workload with goals to build regional capacity to provide climate change information for the region from within the region, to provide much needed climate information in the shortest possible time frame, and to create a platform for sharing the information as widely as possible. Ten years later offers the opportunity for retrospection and evaluation, particularly since a phase 2 initiative is being formulated. By both accident and design, the legacies of the PRECIS-Caribbean initiative include i) the positioning of the Caribbean to pose and answer for itself some of the emerging second-generation climate change questions; ii) the emergence of a regional template for capacity building in the sciences through cooperation; iii) an expanded regional capacity to undertake climate science; and iv) a significant body of climate change and climate science knowledge relevant to and at the scale of the Caribbean region.

CORRESPONDING AUTHOR: Michael A. Taylor, Department of Physics, University of the West Indies, Mona, Jamaica E-mail: michael.taylor@uwimona.edu.jm

By the beginning of the current century, there was heightened recognition that the Caribbean is highly vulnerable to the effects of climate change. Yet, there was very little climate change science information for the region and at the scale of the small islands that make up most of the region. To fill the gap, a group of regional scientists representing three institutions and four territories (Barbados, Belize, Cuba, and Jamaica) initiated a project to provide dynamically downscaled climate change information for the Caribbean. The Providing Regional Climates for Impacts Studies (PRECIS)-Caribbean initiative was premised on a shared workload with goals to build regional capacity to provide climate change information for the region from within the region, to provide much needed climate information in the shortest possible time frame, and to create a platform for sharing the information as widely as possible. Ten years later offers the opportunity for retrospection and evaluation, particularly since a phase 2 initiative is being formulated. By both accident and design, the legacies of the PRECIS-Caribbean initiative include i) the positioning of the Caribbean to pose and answer for itself some of the emerging second-generation climate change questions; ii) the emergence of a regional template for capacity building in the sciences through cooperation; iii) an expanded regional capacity to undertake climate science; and iv) a significant body of climate change and climate science knowledge relevant to and at the scale of the Caribbean region.

CORRESPONDING AUTHOR: Michael A. Taylor, Department of Physics, University of the West Indies, Mona, Jamaica E-mail: michael.taylor@uwimona.edu.jm

INFORMATION DEFICIT.

A regionally formulated and implemented initiative produces dynamically downscaled climate change scenarios for the Caribbean.

By the year 2000, with the help of a swing to a more active phase of hurricane activity, the interest in and awareness of climate change as a developmental issue for the Caribbean region was heightened. The vulnerability of the Caribbean Small Island Developing States (SIDS) to climate variations was becoming evident, accompanied by a growing sentiment that this vulnerability would only be exacerbated under climate change and its warmer temperatures, rising sea levels, changing rainfall regimes, and more intense extreme events. The Caribbean SIDS' vulnerability arises due to a combination of the territories' relative isolation, small landmasses, concentrations of population and infrastructure in coastal areas, limited economic base, and heavy dependency on natural resources. The Caribbean Community Climate Change Centre [CCCCC (5Cs)] was formed in 2004, only a few years later, as testament to the priority that the region's governments were placing on climate change.

Yet, in 2001, very little climate change science information existed for the Caribbean at the scale of the Caribbean. What was known was gleaned from global climate models (GCMs), and it was suggested, for example, that the region, as an averaged whole, would experience a temperature increase of 1°–3°C and more extremes in rainfall by the middle to the end of the century (see, e.g., Wigley and Santer 1993; Singh 1997; McLean and Tysban 2001; Houghton et al. 2001). This kind of information, while useful, was not representative enough for the region's policy and decision makers, who were increasingly demanding finer-scale projections with subregional and, if possible, island-scale differentiation. A number of techniques offered the possibility of obtaining regional climate information. These included using high- and variable-resolution atmosphere–ocean coupled global circulation models (AOGCMs), and employing empirical/statistical and statistical/ dynamical downscaling methods and/or running regional (or nested limited area) climate models. Each method possesses inherent advantages and disadvantages (Mearns et al. 2003).

In September 2003 a group of Caribbean climate scientists met in Havana, Cuba, with the intention of correcting the information imbalance. The consensus was to pursue dynamical downscaling as the means of providing downscaled climate information. The immediate tasks, then, were the acquisition of a regional climate model (RCM) and learning how to run the model. Later tasks would involve learning how to interpret the model output for the value added. The model chosen was the Providing Regional Climates for Impacts Studies (PRECIS) regional model, which had recently been released by the Hadley Centre of the United Kingdom and was being promoted as ideal for meeting the climate information needs of developing countries and small island states (see “About PRECIS” for more information). Funding for the PRECIS workshop in Havana was from the Global Environment Facility (GEF) via the Mainstreaming Adaptation to Climate Change (MACC) Project and the Japanese Trust Fund operated by the World Bank. Coordination of the workshop was by both the MACC Project and the Cuban Meteorological Institute [Instituto de Meterología (INSMET)].

Arising from the workshop was the structure for a multicountry collaborative initiative that could quickly generate the suite of high-resolution climate projections needed, using a subset of the Special Report on Emissions Scenarios (SRES). The projections would initially support two regional projects— the Adaptation to Climate Change in the Caribbean (ACCC Project) based at the time in Barbados and a GEF–United Nations Development Programme (UNDP)-funded project “Capacity building for stage II adaptation to climate change in Central America, Mexico and Cuba,” for which the executing agency was the Water Center for the Humid Tropics of Latin America and the Caribbean (CATHALAC) based in Panama. Implicit also were intentions by the scientists i) to increase the capacity to provide regional science solutions to a regional science problem by further developing the capacity to do modeling within the region and ii) to gain an appreciation for, and later contribute to, the science of regional modeling, using the Caribbean and its small islands as the basis for doing so.

THE PRECIS-CARIBBEAN AGENDA.

The outcome of the workshop was the decision to install the PRECIS model at three regional institutes spanning five countries. By so doing, the region could quickly produce downscaled results through a shared workload. The institutions included the three campuses of the University of the West Indies (UWI) located in Barbados, Jamaica, and Trinidad; the INSMET in Cuba; and the 5Cs in Belize. Within a few months, Trinidad withdrew because of human capacity limitations. Two domains were defined to take advantage of PRECIS's two available resolutions—a Caribbean domain at 50-km resolution encompassing the area from the Bahamas to Guyana and Central America (UWI-Jamaica and INSMET to perform experiments), and a smaller eastern Caribbean domain at 25-km resolution to capture the smaller islands in the archipelago (UWI-Barbados to perform experiments). A western Caribbean domain at 25 km was later added (see Fig. 1). The model was to be forced by the Third Hadley Centre Atmosphere Model version H [(HadAM3H) and later by HadAM4 version P (HadAM3P)] (Jones et al. 2003) and ECHAM4 global models, and the experiments done using a time-slice approach simulating present (1960–90) and end-of-century (2071–99) climate for the A2 and B2 SRES scenarios (see “SRES” for more information). Table 1 details the initial division of runs.

Fig. 1.
Fig. 1.

(a) Caribbean domains—Big domain at 50 km (all Caribbean, Central America, southern United States, and northern South America) and two smaller domains at 25 km: western Caribbean (red) and eastern Caribbean (green). Also shown are the land–sea mask of the (b) HadAM3H GCM and (c) PRECIS RCM for the big domain, and (d) topography for an 18°N cross section through the big domain as represented by Global 30-arc-second elevation dataset (GTOPO 30) quasi observations (green), HadAM3H (red), and PRECIS (black).

Citation: Bulletin of the American Meteorological Society 94, 7; 10.1175/BAMS-D-11-00235.1

Table 1.

Initial division of runs.

Table 1.

ABOUT PRECIS

  • Pronounced PRAY-SEE.
  • Means Providing Regional Climates for Impacts Studies.
  • Is an atmospheric and land surface model of limited area and high resolution that is locatable over any part of the globe.
  • Has a horizontal resolution of 0.44° (~50 km) or 0.22° (~25 km) and 19 levels in the vertical.
  • Formulates the following processes: dynamical flow, atmospheric sulphur cycle, clouds and precipitation, radiative processes, land surface, and deep soil.
  • Forced at its lateral boundaries by the simulations of a high-resolution global model, for example, HadAM3H and ECHAM4.
  • Built by the Hadley Centre but run locally.
  • Works on a Linux-based personal computer (PC).
  • Available online at www.metoffice.gov.uk/precis/.

There were no shortage of problems, and the ambitious target of producing a full suite of runs in two years utilizing all available scenarios for both resolutions and domains quickly proved unrealistic. In the first instance, the available hardware at each site required upgrading. There were also delays in acquiring the boundary conditions, and the data storage problem seemed insurmountable. For the latter problem, a tape drive solution was initially attempted but was soon abandoned for hard drives. It was the hard drives that were eventually exchanged between institutes when runs were finally completed three years later.

In 2006, a follow-up workshop sponsored by the 5Cs took place in Belize City, Belize. The first three days of the workshop were for new users of PRECIS and the last two days were for sharing results from the first allocation of runs. In 2007 a sensitization document about the project entitled “Glimpses of the Future” (Taylor et al. 2007) was produced that offered initial results for use by policy makers. A website for distribution of the data was also launched (see http://precis.insmet.cu/Precis-Caribe.htm). By 2008 new model runs were added to fill in the intermediate periods not covered by the time-slice approach. By 2010 papers began appearing in peer-reviewed journals using the downscaled model results (e.g., Campbell et al. 2011; Charlery and Nurse 2010; Taylor et al. 2011, 2013).

AN EMERGING PICTURE—SOME RESULTS.

Validation.

The first set of results focused on assessing how well the model reproduced the present climate. Validation is necessary, since it provides both insight into the model's skill and the basis for user confidence in the model's results, particularly when the results are to be used in climate impact assessment studies. Various issues complicate the model validation process, for example, the difficulty in identifying model error because of the complex and sometimes poorly understood nature of the climate itself, and the lack of reliable and consistent observations, which is a real problem in the Caribbean.

SRES

The scenarios used in the modeling are the storylines of future global development established by the IPCC and reported on in the SRES (Nakićenović et al. 2000). The SRES scenarios quantify how greenhouse gas emissions could change over the twenty-first century in the absence of policy interventions to reduce the emissions. In all there are 40 different scenarios divided into four families (A1, A2, B1, and B2), each with an accompanying storyline that describes the relationships between future greenhouse gas emission levels and driving forces such as demographic, social and economic, and technological developments. The families represent a range of equally plausible futures—from low-emission to high-emission futures. The A family, or high-emissions scenarios, describes a future world of very rapid economic growth, global population that peaks in midcentury and declines thereafter, with the rapid introduction of new and more efficient technologies. The A2 storyline describes a very heterogeneous world. The underlying theme is self-reliance and preservation of local identities. The B family describes relatively low-emissions scenarios. The B2 storyline describes a world in which the emphasis is on local solutions to economic, social, and environmental sustainability.

Some validation results were offered by Campbell et al. (2011), who compared PRECIS's modeled patterns of temperature and precipitation with reanalysis datasets and available observations. They showed the mean Caribbean climatologies to be generally captured by the model, with the relative timing of temperature and precipitation maxima and minima being reproduced. This included the model's reproduction of the Caribbean midsummer rainfall minimum, which is a significant feature of many of the larger Caribbean islands (Curtis and Gamble 2008; Magaña et al. 1999). There was, however, also a general underestimation of rainfall amounts across the main Caribbean basin during the wet season and a simulation using temperatures that were too warm over the Caribbean islands but too cold over Central America and northern South America [see Figs. 2b and 2d, and Figs. 2 and 3 from Campbell et al. (2011)]. Other more recent validation efforts focus on the model's ability to reproduce other key features of Caribbean climate, for example, the Caribbean low-level jet (see, e.g., Taylor et al. 2011, 2013).

Fig. 2.
Fig. 2.

Mean seasonal differences [model – Climatic Research Unit (CRU )] in (a),(b) temperature (°C) and (c),(d) precipitation (%) with respect to the (CRU ) database (Mitchell et al. 2004) for the period of May–Oct 1961–89. Data for (a),(c) HadAM3P and (b),(d) PRECIS are interpolated to the CRU 0.5° grid.

Citation: Bulletin of the American Meteorological Society 94, 7; 10.1175/BAMS-D-11-00235.1

Fig. 3.
Fig. 3.

Mean changes in the annual mean 1.5-m temperature for 2071–99 with respect to 1961–89 as simulated by (left) HadAM3P and (right) PRECIS for the SRES A2 scenario.

Citation: Bulletin of the American Meteorological Society 94, 7; 10.1175/BAMS-D-11-00235.1

Over some regions of the Caribbean domain, the temperature and rainfall biases with respect to observation or reanalysis datasets were notably smaller in the RCM simulation in comparison to the driving GCM—for example, the smaller temperature biases over Cuba and Jamaica in Fig. 2b versus Fig. 2a. It was also noted that uncertainties in the RCM fields may arise in part from the biases of the driving GCM—for example, the dry bias of, in particular, the eastern Caribbean seen in both RCM and GCM simulations (Fig. 2). These observations speak to the larger question of the value added for the region by doing dynamical downscaling versus simply utilizing available information from, for example, the forcing GCM simulations. The question of added value is a pertinent one in the context of the small island states. Though not the initial focus of the PRECIS-Caribbean initiative and so not addressed in the initial papers arising from the initiative, the question is being examined in forthcoming studies (e.g., A. Centella et al. 2012; personal communication; Karmalkar et al. 2013). Campbell et al. (2011), however, note that if nothing else, there is benefit to be had from the better representation of the smaller island landmasses and the increase in topographic and coastal details afforded by PRECIS (see Figs. 1b–d). The land–sea thermal contrasts and topography produce more rain over the Caribbean islands, which offsets somewhat the influence of the dry bias of the driving GCM (Campbell et al. 2011).

Future changes.

Some of the first future climate change projections for the Caribbean from a regional model were documented by Taylor et al. (2007) on the basis of the first round of PRECIS simulations driven by the HadAM3P GCM. They showed a Caribbean that is 1°–5°C warmer in the annual mean by the 2080s (a 30-yr period from 2071 to 2100), and one also characterized by i) greater warming in the northwest (Jamaica, Cuba, Hispaniola, and Belize) in comparison to the eastern Caribbean island chain and ii) greater warming in the summer months than in the drier early months of the year.

Figure 3 shows the projected end-of-century temperature change for the larger domain from both the driving GCM and the RCM for the A2 scenario. The temperature change for the Caribbean islands is both larger than and distinguishable from that for the surrounding sea in the RCM, as evident for Jamaica, Cuba, and Hispaniola. Centella et al.'s (2008) report projected annual warming of the order of 4.5°C for the larger Caribbean islands in the RCM, which is larger than both the projected increase in the GCM (~3°C) and the projected increase over the Caribbean Sea in the RCM (~2.9°C). Similarly, temperature changes over the smaller islands represented in the RCM (but not in the GCM) are distinguishable from the temperature of the surrounding ocean (Karmalkar et al. 2013). This is likely due to evaporative heat loss being less over the land and the greater thermal inertia of the oceans (Houghton et al. 2001).

The patterns of change in annual rainfall corroborate a future drying trend suggested by the GCMs but make clearer some of the possible sub-regional variations. Figure 4 captures some of the main results. For the A2 scenario, the PRECIS simulations suggest a significant reduction of mean annual rainfall (10%–50%) by the end of the century i) between 10° and 24°N—that is, the latitudinal band encompassing the Caribbean Sea and most of Central America; and ii) over northeastern South America. Above 24°N, a 10%–30% increase in precipitation is projected. Increased precipitation is also projected for Costa Rica and Panama (in the case of Panama, the change is different from the driving model) and parts of Colombia. For the B2 scenario, a significant difference is that the wet area in the north extends southward to 21°N, so that rainfall increases also occur over Cuba.

Fig. 4.
Fig. 4.

Annual mean changes in precipitation (%) for 2071–99 with respect to 1961–89 as simulated by (left) HadAM3P and (right) PRECIS. Top (bottom) shows results for the B2 (A2) emission scenario.

Citation: Bulletin of the American Meteorological Society 94, 7; 10.1175/BAMS-D-11-00235.1

The analyses of Taylor et al. (2011), Campbell et al. (2011), and Centella et al. (2008) provide much more detail, particularly with respect to differences across the wet and dry seasons. For example, Centella et al. (2008) and Campbell et al. (2011) show that in the mean wet season (May–November), the areas with diminished rainfall occupy almost the entire PRECIS domain (i.e., no change in sign north and south of 24°N), with the largest change occurring over the Caribbean Sea, near south of Cuba, and over the island of Hispaniola (30%–50% drier). Centella et al. (2008) also analyze the ECHAM-forced simulations, that is, an expanded ensemble of projections. They note variations in the results between the two (ECHAM and Hadley) GCM-forced experiments. For example, for the ECHAM-only projections of annual mean end-of-century rainfall, the drying pattern over the Caribbean Sea and parts of Central America extends also over the Gulf of Mexico, and the wet areas over the northern part of South America are more pronounced and wider in extent (Centella et al. 2008). They also note i) generally larger reductions in endof- century rainfall amounts for the Hadley GCM– forced simulations than for the ECHAM runs; ii) a less pronounced pattern of drying over the Caribbean Sea for the B2 scenario, irrespective of the driving model; and iii) major differences in the dry season, as PRECIS_Hadley tends to maintain the latitudinal spatial gradient with a dry band over the Caribbean Sea (10% and lower) and a wet area above 21°N (10%–70%) in comparison to PRECIS_ECHAM, which produces the driest zones over the northwestern Caribbean, Gulf of Mexico (10%–50%), and the southern areas of the eastern Caribbean (10%–30%).

Figure 5 shows a consensus diagram for future precipitation projections for an ensemble of four realizations (the two emission scenarios for each of the two GCM forcings). The diagram suggests a high level of agreement that mean annual rainfall will be substantially reduced over large areas of the Caribbean Sea and Central America, mainly associated with a reduction of precipitation in the wet season. The latter is in general agreement with the multimodel regional projection reported in the Intergovernmental Panel on Climate Change (IPCC) Fourth Assessment Report (AR4) (Solomon et al. 2007) but with additional spatial detail. For the Caribbean, which is so dependent on tropical convective processes, Fig. 5 raises the question of why a warmer future world, with warmer tropical ocean temperatures, should yield such a robust picture of drying. The usefulness of the PRECIS-Caribbean initiative is that it not only contributes to the formulation of such context- specific valid research questions but also provides data that can help facilitate the answering of the same questions (see, e.g., Taylor et al. 2013).

Fig. 5.
Fig. 5.

Number of simulations projecting precipitation increase for the 2080s (left) annually, and during the (middle) wet season and (right) dry season.

Citation: Bulletin of the American Meteorological Society 94, 7; 10.1175/BAMS-D-11-00235.1

LESSONS AND LEGACY.

In retrospect, the PRECIS-Caribbean initiative offers lessons to both onlookers and those who were directly involved, and leaves a legacy upon which future scientific efforts can be built. The lessons include:

  • i) The value of collaboration. At the initiation of the project, the ability of any one of the participating institutions to deliver the desired range of model simulations (spanning scenarios, resolutions and forcing GCMs) was limited. It was the ability to undertake parallel runs in multiple institutions, as well as the opportunities for mutual exchange with respect to troubleshooting and encouragement, and a shared burden for the importance of the task that allowed for the “timely” delivery of downscaled climate information for the Caribbean region. The success of the collaborative approach points to a modality for maximizing the use of limited regional resources to produce regionally relevant science information without the need for replication in individual territories.
  • ii) The value of institutional will. Available funding for the respective institutions to undertake the project was uneven and in most cases limited to monies provided for initial hardware and the attendance at meetings. The pursuance of the assigned tasks (and more) and the persistence beyond the initially anticipated lifetime of the project were largely driven by project buy-in, a shared understanding of the significance of the work being undertaken, and the necessity of having the data generated within the region for the region—that is, lest the region be left behind in global efforts to define contextual vulnerabilities and to develop targeted resilience options.

Now approaching a decade since its inception, the legacies of the PRECIS-Caribbean project are more apparent and include:

  • i) Data: A significant amount of future climate data for the Caribbean at scales closer to Caribbean island scales has been generated. All the data are available through the PRECIS-Caribe website hosted by INSMET and are also available from the climate change information clearinghouse hosted by the 5Cs.
  • ii) Capacity: The capacity to confidently do modeling work within the region and to apply downscaling techniques has been enhanced at each of the original centers involved. The knowledge gained by the lead researchers has been passed on to graduate students and other technical staff. There has also been the addition of a fifth modeling center at the Anton de Kom University of Suriname. Modeling at Anton de Kom was initiated using regional expertise garnered under the PRECIS-Caribbean initiative and concentrates on climate change scenarios for the Caribbean states in northern South America.
  • iii) Growing body of research: There is growing evidence of cross-disciplinary research examining Caribbean climate change questions emanating from within the region and using the PRECIS project data. These include attempts to better define each territory's vulnerabilities and to use evidence-based research as the impetus for incorporating climate change in regional and national policies for planning and development. The data have been used by at least eight countries to date in their Second National Communication reporting requirements for the United Nations Framework Convention on Climate Change (UNFCCC). The data have also been used for studies of the economic cost of climate change for specific sectors in selected Caribbean countries (ECLAC 2012), in science reviews (e.g., Campbell et al. 2011; Charlery and Nurse 2010; Taylor et al. 2011), and in the examinations of the impact of climate change on water resources (Cashman et al. 2010) and fisheries (Nurse 2011).
  • iv) Science: The context of largely small islands and the lack of information at that scale drove the initial mandate for producing downscaled climate information. The value added by downscaling to 25 and 50 km, with prospects for even smaller scales, is, on the one hand, seen through the ability to produce the climates of coastlines and topography, which are so important to the region. The value over and beyond better horizontal resolution is, however, to be investigated. Questions to be investigated include: Is the additional magnitude of projected change (over, e.g., GCM results) justifiable, especially in regions where they reverse the sign of the driving GCM? Are the RCM results comparable to that from other downscaling efforts, for example, statistical downscaling? What proportion of the biases derives from the driving GCM, or from the model physics or model configuration? For example, to what extent is the severe warming observed over the Amazon in Fig. 3 related to uncertainties in the land surface model behavior in the parent model (see, e.g., Cox et al. 2004)? What is the impact of domain size and location on the projected results? The information generated and the skills gained in pursuing the PRECIS-Caribbean initiative provide the backdrop for examining these science questions related to scale as well as other questions that simply emerge on the basis of the resulting future picture. Such examinations are underway. Charlery and Nurse (2010) is an example of the region's attempt to develop new techniques for further downscaling to address the small islands' sizes (including subisland scales, e.g., watersheds) and data limitations.
  • v) A future agenda: A Caribbean climate modelers group has emerged from the PRECIS-Caribbean project (Table 2). The group has met annually for the last three years, and has defined and articulated priority actions to meet regional modeling needs and a coordinated collaborative approach to doing so. The group is actively seeking funding for its proposals. In the meantime the group has also begun phase 2 of the PRECIS project involving an ensemble of perturbation experiments and begun simulations with other regional models, for example, the Weather Research and Forecasting (WRF) model. It is also positioning the region to participate in the Coordinated Regional Downscaling Experiment (CORDEX) project and to generate downscaled scenarios premised on the new response concentration pathways.

The PRECIS-Caribbean initiative, notwithstanding its challenges, has been instrumental in moving the Caribbean region from a position of limited or no representation in the initial IPCC reports and very little climate scenario data at the scale of the majority of the region to one where downscaled data exist, which is spurring research and beginning to influence regional policy and development plans. The PRECIS-Caribbean initiative is an example of what can be achieved through collaboration and collective will, and the shared desire to see the advancement of a region without waiting on external factors to occasion it.

Table 2.

Caribbean modeling group.

Table 2.

Acknowledgments

The Caribbean modelers group acknowledges the invaluable contributions of the Hadley Centre, United Kingdom, and the Caribbean Community Climate Change Centre (5Cs). Without the assistance of both institutions, the PRECIS-Caribbean project would not have been possible.

REFERENCES

  • Campbell, J. D., , M. A. Taylor, , T. S. Stephenson, , R. A. Watson & , and F. S. Whyte, 2011: Future climate of the Caribbean from a regional climate model. Int. J. Climatol., 31, 18661878, doi:10.1002/joc.2200.

    • Search Google Scholar
    • Export Citation
  • Cashman, A., , L. Nurse & , and J. Charlery, 2010: Climate change in the Caribbean: The water management implications. J. Environ. Dev., 19, 4267.

    • Search Google Scholar
    • Export Citation
  • Centella, A., , A. Bezanilla & , and K. Leslie, 2008: A study of the uncertainty in future Caribbean climate using the PRECIS regional climate model. Community Caribbean Climate Change Centre Tech. Rep., 16 pp.

    • Search Google Scholar
    • Export Citation
  • Charlery, J. & , and L. Nurse, 2010: Areal downscaling of global climate models: An approach that avoids data remodeling. Climate Res., 43, 241249, doi:10.3354/ cr00875.

    • Search Google Scholar
    • Export Citation
  • Cox, P. M., , R. A. Betts, , M. Collins, , P. P. Harris, , C. Huntingford & , and C. D. Jones, 2004: Amazonian forest dieback under climate-carbon cycle projections for the 21st century. Theor. Appl. Climatol., 78, 137156.

    • Search Google Scholar
    • Export Citation
  • Curtis, S. & , and D. W. Gamble, 2008: Regional variations of the Caribbean mid-summer drought. Theor. Appl. Climatol., 94, 2534, doi:10.1007/s00704-007-0342-0.

    • Search Google Scholar
    • Export Citation
  • ECLAC, cited 2012: Review of the economics of climate change in the Carribean. [Available online at www.eclac.cl/cgi-bin/getprod.asp?xml=/portofspain/noticias/paginas/4/34444/P34444.xml&xsl=/portofspain/tpl-i/p18f.xsl&base=/portofspain/tpl-i/top-bottom.xsl.]

    • Search Google Scholar
    • Export Citation
  • Houghton, J. T., , Y. Ding, , D. J. Griggs, , M. Noguer, , P. J. van der Linden, , X. Dai, , K. Maskell & , and C. A. Johnson, Eds., 2001: Climate Change 2001: The Scientific Basis. Cambridge University Press, 881 pp.

    • Search Google Scholar
    • Export Citation
  • Jones, R. G., , M. Noguer, , D. Hassell, , D. Hudson, , S. Wilson, , G. Jenkins & , and J. Mitchell, 2003: Workbook on generating high resolution climate change scenarios using PRECIS. UNDP, GEF, and Met Office Hadley Centre Manual, 32 pp.

    • Search Google Scholar
    • Export Citation
  • Karmalkar, A. V., , M. A. Taylor, , J. Campbel l, , T. Stephenson, , M. New, , A. Centella, , A. Bezanilla & , and J. Charlery, 2013: A review of observed and projected changes in climate for the islands in the Caribbean. Atmósfera, 26, 283309.

    • Search Google Scholar
    • Export Citation
  • Magaña, V., , J. A. Amador & , and S. Medina, 1999: The midsummer drought over Mexico and Central America. J. Climate, 12, 15771588.

  • McClean, R. F. & , and A. Tysban, 2001: Coastal zones and marine ecosystems. Climate Change 2001: Impacts, Adaptation, and Vulnerability, J. J. McCarthy et al., Eds., Cambridge University Press, 343379.

    • Search Google Scholar
    • Export Citation
  • Mearns, L. O., , F. Giogi, , P. Whetton, , D. Pabon, , M. Hulme & , and M. Lal, 2003: Guidelines for the use of climate scenarios developed from regional climate model experiments. IPCC and TGCIA, 38 pp. [Available online at www.ipcc-data.org/guidelines/dgm_no1_v1_10-2003.pdf.]

    • Search Google Scholar
    • Export Citation
  • Mitchell, T. D., , T. R. Carter, , P. D. Jones, , M. Hulme & , and M. New, 2004: A comprehensive set of high-resolution grids of monthly climate for Europe and the globe: The observed record (1901–2000) and 16 scenarios (2001–2100). Tyndall Centre for Climate Change Research Working Paper 55, 30 pp.

    • Search Google Scholar
    • Export Citation
  • Nakićenović, N. & , and R. Swart, Eds., 2000: Special Report on Emissions Scenarios. Cambridge University Press, 599 pp.

  • Nurse, L., 2011: The implications of global climate change for fisheries management in the Caribbean. Climate Dev., 3, 228241.

  • Singh, B., 1997: Climate changes in the greater and southern Caribbean. Int. J. Climatol., 17, 10931114.

  • Solomon, S., , D. Qin, , M. Manning, , M. Marquis, , K. Averyt, , M. M. B. Tignor, , H. L. Miller Jr. & , and Z. Chen, Eds., 2007: Climate Change 2007: The Physical Science Basis. Cambridge University Press, 996 pp.

    • Search Google Scholar
    • Export Citation
  • Taylor, M. A., and Coauthors, 2007: Glimpses of the future: A briefing from the PRECIS Caribbean Climate Change Project. Caribbean Community Climate Change Centre, 24 pp.

    • Search Google Scholar
    • Export Citation
  • Taylor, M. A., , T. S. Stephenson, , A. Owino, , A. A. Chen & , and J. D. Campbell, 2011: Tropical gradient influences on Caribbean rainfall. J. Geophys. Res., 116, D00Q08, doi:10.1029/2010JD015580.

    • Search Google Scholar
    • Export Citation
  • Taylor, M. A., , F. S. Whyte, , T. S. Stephenson & , and J. D. Campbell, 2013: Why dry? Investigating the future evolution of the Caribbean low level jet to explain projected Caribbean drying. Int. J. Climatol., 33, 784792, doi:10.1002/joc.3461.

    • Search Google Scholar
    • Export Citation
  • Wigley, T. M. L. & , and B. D. Santer, 1993: Future climate of the Gulf/Caribbean basin from atmospheric general circulation models. Climatic Change in the Intra-Americas Sea, G. A. Maul, Ed., Chapman and Hall, 3154.

    • Search Google Scholar
    • Export Citation
Save