Climate forecasts advances in the Caribbean DR CEDRIC

Climate forecasts – advances in the Caribbean DR. CEDRIC VAN MEERBEECK Climatologist cmeerbeeck@cimh. edu. bb WAZITA SCOTT, MSC Assistant Climatologist wscott@cimh. edu. bb CARIBBEAN INSTITUTE FOR METEOROLOGY AND HYDROLOGY Climate Product and Climate Services Toolkit Workshop – St. Lucia 8 -17 January 2018, Vieux Fort, St. Lucia

Climate Forecasts – what and why?

Concern. . . 3 Credit: WMO, 2013

Science and Policy: Information Chains Example: Information Chain in Climate Risk Management SCIENCE Climate Science Climate Information Products POLICY Climate Information and Products for Disasters National Emergency System Early Warning Information Systems (EWIS) New Research Questions When links are not present, create them! When Chains are established and strong we do not need to “market” 4 Adapted from: Roger Pulwarty, NOAA Local Implementation New Knowledge Demands

Climate Prediction Framework state of the atmosphere Short-range (weather) forecasts 5 Adapted from: GFCS

Climate Prediction Framework Mid-range (weather) forecasts state of the atmosphere Short-range (weather) forecasts 6 Adapted from: GFCS

Climate Prediction Framework Long-range (i. e. seasonal) forecasts e. g. sea surface temperatures Mid-range (weather) forecasts state of the atmosphere Short-range (weather) forecasts 7 Adapted from: GFCS

Climate Prediction Framework Climate change projections / scenarios Long-range (i. e. seasonal) forecasts e. g. CO 2 emissions e. g. sea surface temperatures Mid-range (weather) forecasts state of the atmosphere Short-range (weather) forecasts 8 Adapted from: GFCS

Decision-making across timescales Begin monitoring mid-range and short-range forecasts Update contingency plans Train volunteers Sensitize community Enable early-warning system Continue monitoring shorter-time-scale forecasts Mobilize assessment team Alert volunteers Warn community Local preparation activities Hellmuth et al 2011 Deploy assessment team Activate volunteers Evacuate community

Climate Forecasts – how and why? A recap after Tony Barnston (IRI)

Climate forecasts need to be expressed probabilistically, because …

Climate forecasts need to be expressed probabilistically, because there is usually a wide distribution of possibilities even with our better-thance accuracy.

What probabilistic forecasts represent Near-Normal Below Normal Above Normal FREQUENCY Historical distribution (climatological distribution) (33. 3%, 33. 3%) Forecast distribution (15%, 32%, 53%) NORMALIZED RAINFALL Historically, the probabilities of above and below are 0. 33. Shifting the mean by half a standarddeviation and reducing the variance by 20% changes the probability of below to 0. 15 and of above to 0. 53. (Courtesy Mike Tippett)

Abbreviating a predicted shift in the probability distribution: Terciles (Below normal, near normal, above normal) Climatological probabilities = 1/3 33% Below| Data: 33% Near | Above | || ||||. | || | | | 0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 200 Rainfall Amount (mm) (30 years of historical data for one station and season)

Example of a climate forecast with a strong probability shift 10% Below| 25% Near | 65% Above 0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 200 Rainfall Amount (mm)

Example of a climate forecast with a weak probability shift 25% Below| 35% Near | 40% Above 0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 200 Rainfall Amount (mm)

Example of a climate forecast with no probability shift 33% Below| 33% Near | 33% Above 0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 200 Rainfall Amount (mm)

Climate Forecasts – based on what?

During El Niño, autumn and winter rainfall usually higher in the Bahamas, but in over the eastern and southern Caribbean. La Niña tends to result in the opposite rainfall gradient. E. Caribbean: ’ 09 -’ 10 El Niño - record dry; ‘ 10 -’ 11 La Niña - record wet. NOAA CPC – ENSO monitoring EL NINO LA NINA

Forecast warm North Atlantic and Caribbean Sea SSTs for March to June 2013 forecast of likely above- to near-normal MJJ 2013 rainfall for the Lesser Antilles. El Niño region Caribbean & tropical Atlantic

Forecast warm North Atlantic and Caribbean Sea SSTs for March to June 2013 forecast of likely above- to near-normal MJJ 2013 rainfall for the Lesser Antilles. + Other factors make rainfall (and temperature) predictable at seasonal timescale… How predictable? à Rainfall totals over the next 3 months: 30% better than guessing à Average temperature over the next 3 months: up to 70% better than guessing

Cari. COF climate outlooks – how to …?

Regional Climate Outlook Forums Cari. COF - FCAC Cari. COF FCAC

General methodology I

General methodology II

General methodology III

Cari. COF climate outlooks – how to in CPT?

CPT probabilistic April – May – June (AMJ) rainfall forecast ENSO (seen in tropical Pacific Sea Surface Temperature (SST) anomalies) and Caribbean/tropical North Atlantic SSTs are the most important drivers of seasonal rainfall variability. Our forecast experiments take Pacific and Caribbean/tropical North Atlantic SSTs into account by building ensembles of regression models (Canonical Correlation Analysis - CCA). CCA experiments using CPT: 1) Predictor is observed SST over the tropical Atlantic and Pacific over February (data source: NOAA ERSSTv 3 b, obtained from the IRI data library) 2) Predictor is observed SST over the tropical North Atlantic over February 3) Predictor is predicted SST over the tropical Atlantic and Pacific over AMJ (data source: NOAA CPC CFSv 2, March initialisation) 4) Predictor is predicted SST over the tropical North Atlantic 5) Predictor is predicted rainfall totals over the Caribbean (data source: ECHAM 4. 5 ensemble 24, obtained from the IRI data library)

CPT experiment 1 --- April to June rainfall forecast Data: Predictor = February SST observations 1982 -2015 --- 140 E-20 W and 30 N-20 S (i. e. Tropical Pacific & Atlantic) [Source: ERSSTv 3 b, from IRI data library] Predictand = AMJ rainfall for 474 Caribbean stations 1982 -2014 Parameters: CCA modes: max = 5; optimum = 4 X modes: max = 8; optimum = 4 Y modes: max = 8; optimum = 4 Missing Y data threshold: 20% missing values, 10% missing stations, leading to 428 stations included in hindcast and forecast Goodness index (transformed Y) = !!0. 095 Very Limited!!

CPT Experiment 1 - AMJ 2015 rainfall forecast

CPT Experiments ensemble - AMJ 2015 rainfall forecast CPT ENSEMBLE MEAN FORECAST

Supporting national rainfall forecasts CARICOF national/sub-regional objective forecasts & outlook guidance: 19) Cuba - West: Cuba – Central: Cuba – East: AMJ JAS A = 32% ; N = 40% ; B = 28% NATIONAL TECHNICAL A = 25% ; N = 25% ; B = 50% OUTLOOK TEMPLATES A = 42% ; N = 26% ; B = 32% A = 19% ; N = 30% ; B = 51% A = 38% ; N = 20% ; B = 42% A = 18% ; N = 40% ; B = 42% Note: CPT CCA with ERSSTv 3 b and CFSv 2 SSTs showed negative skill with no to poor discrimination for AMJ and; moderate to fair skill with poor to fair discrimination for JAS. 20) Guadeloupe: Martinique: French Guiana: French Northern Antilles: AMJ AMJ A = 34% ; N = 34% ; B = 32% A = 33% ; N = 35% ; B = 32% A = 33% ; N = 34% ; B = 33% A = 34% ; N = 34% ; B = 32% Note: For the Lesser Antilles there is an agreement between several dynamical models (EUROSIP, JMA) ) with below or normal precipitation over the area. Statistic SST/RR models (with some skill) show a shift to AN over Martinique, Guadeloupe and the Iles du Nord for AMJ 2015. (predictor SST over tropical North Atlantic and Pacific, with positive ENSO anomaly and negative TNA). Over French Guiana : The statistic model has low skill and suggests a shift to AN. The dynamical models suggest a shift to BN. Normal to wet over Guadeloupe, except for Marie Galante with dry conditions.

Cari. COF CPT ensemble forecast & National/sub-regional rainfall outlooks

CCA experiments – latest methodological update example from zero month lead time (in this case: Sep-Oct-Nov precip forecast) 1) Predictor is observed SST over the tropical Atlantic and Pacific over July (data source: NOAA ERSSTv 3 b, obtained from the IRI data library) 2) Predictor is observed SST over the tropical North Atlantic over July 3) Predictor is predicted SST over the tropical Atlantic and Pacific over SON (data source: NOAA CPC CFSv 2, August initialisation) 4) Predictor is predicted CFSv 2 SST over the tropical North Atlantic 5) Predictor is predicted SST over the tropical Atlantic and Pacific over SON (data source: NMME multi-model ensemble mean, August initialisation) 6) Predictor is predicted NMME SST over the tropical North Atlantic For three month lead time forecasts, a similar set of experiments are run. ERSST same month as 0 -month lead, predicted SST covering period of forecast. Besides the major control of ENSO (here represented by Pacific tropical SST anomalies) and tropical north Atlantic SSTs on Caribbean rainfall variability, these experiments take the contrast between Pacific and Caribbean/trop. N Atlantic SSTs into account, as those factors are regarded as the most important drivers of rainfall throughout the Caribbean.

Cari. COF Objective Rainfall Outlook for the Caribbean (National + Regional Ensemble) January-February-March 2018 caricof@cimh. edu. bb

Probabilistic JFM rainfall forecast map caricof@cimh. edu. bb

Climate Forecasts – how to read?

Seasonal Rainfall Outlook III KEY: messages in those rainfall (temperature) forecasts? Forecast: AN 70% NN 20% BN 10% Forecast: AN 50% NN 35% BN 15% Forecast: AN 25% NN 35% BN 40% Forecast: AN 35% NN 35% BN 30% Forecast: AN 33% NN 33% BN 33%

Seasonal Rainfall Outlook III KEY: messages in those rainfall (temperature) forecasts? Forecast: AN 70% NN 20% BN 10% Expect wetter (warmer) than usual for this period (with only a 1 o% chance for drier (cooler)). Very high chance for extremely wet (warm) conditions. This chance is much higher than in most other years Forecast: AN 50% NN 35% BN 15% Forecast: AN 25% NN 35% BN 40% Forecast: AN 35% NN 35% BN 30% Forecast: AN 33% NN 33% BN 33%

Seasonal Rainfall Outlook III KEY: messages in those rainfall (temperature) forecasts? Forecast: AN 70% BN 10% Expect wetter (warmer) than usual for this period (with only a 1 o% chance for drier (cooler)). Very high chance for extremely wet (warm) conditions. This chance is much higher than in most other years Forecast: AN 50% NN 20% NN 35% BN 15% Expect wetter (warmer) than to pretty much like usual for this period, with a confidence of 85%. For every two times we issue this forecast, one time will be wetter (warmer) than usual. Reasonable chance for extremely wet (warm) conditions. This chance is higher than in most other years. Forecast: AN 25% NN 35% BN 40% Forecast: AN 35% NN 35% BN 30% Forecast: AN 33% NN 33% BN 33%

Seasonal Rainfall Outlook III KEY: messages in those rainfall (temperature) forecasts? Forecast: AN 50% BN 10% NN 35% BN 15% Expect wetter (warmer) than to pretty much like usual for this period, with a confidence of 85%. For every two times we issue this forecast, one time will be wetter (warmer) than usual. Reasonable chance for extremely wet (warm) conditions. This chance is higher than in most other years. Forecast: AN 25% NN 20% Expect wetter (warmer) than usual for this period (with only a 1 o% chance for drier (cooler)). Very high chance for extremely wet (warm) conditions. This chance is much higher than in most other years Forecast: AN 70% NN 35% BN 40% Expect drier (cooler) to pretty much like usual for this period, with a confidence of 75%. Low chance for extremely dry (cool) conditions. This chance is still a bit higher than usual. Forecast: AN 35% NN 35% BN 30% Forecast: AN 33% NN 33% BN 33%

Seasonal Rainfall Outlook III KEY: messages in those rainfall (temperature) forecasts? Forecast: AN 50% AN 25% BN 15% NN 35% BN 40% NN 35% BN 30% Little can be said at this time other than that we have a 70% confidence it will be wetter (warmer) or similar to usual for this period. The chance for extremely wet (warm) or dry (cool) is low. Forecast: NN 35% Expect drier (cooler) to pretty much like usual for this period, with a confidence of 75%. Low chance for extremely dry (cool) conditions. This chance is still a bit higher than usual. Forecast: AN 35% BN 10% Expect wetter (warmer) than to pretty much like usual for this period, with a confidence of 85%. For every two times we issue this forecast, one time will be wetter (warmer) than usual. Reasonable chance for extremely wet (warm) conditions. This chance is higher than in most other years. Forecast: NN 20% Expect wetter (warmer) than usual for this period (with only a 1 o% chance for drier (cooler)). Very high chance for extremely wet (warm) conditions. This chance is much higher than in most other years Forecast: AN 70% AN 33% NN 33% BN 33%

Seasonal Rainfall Outlook III KEY: messages in those rainfall (temperature) forecasts? Forecast: AN 50% AN 25% AN 35% BN 15% NN 35% BN 40% NN 35% BN 30% Little can be said at this time other than that we have a 70% confidence it will be wetter (warmer) or similar to usual for this period. The chance for extremely wet (warm) or dry (cool) is low. Forecast: AN 33% NN 35% Expect drier (cooler) to pretty much like usual for this period, with a confidence of 75%. Low chance for extremely dry (cool) conditions. This chance is still a bit higher than usual. Forecast: BN 10% Expect wetter (warmer) than to pretty much like usual for this period, with a confidence of 85%. For every two times we issue this forecast, one time will be wetter (warmer) than usual. Reasonable chance for extremely wet (warm) conditions. This chance is higher than in most other years. Forecast: NN 20% Expect wetter (warmer) than usual for this period (with only a 1 o% chance for drier (cooler)). Very high chance for extremely wet (warm) conditions. This chance is much higher than in most other years Forecast: AN 70% NN 33% BN 33% We are confident that little can be said at this time, BUT The chance for extremely wet (warm) or dry (cool) is very low and lower than usual.

TIME FOR A BREAK!

Climate Forecasts – what’s the value?

Forecast Value Forecast quality • Do the forecasts correspond well with the observations? Forecast value • Do the forecasts enable beneficial decisions? – Simplistic games such as weather roulette shows that pre-positioning resources proportionate to forecast probabilities leads to profits. – Thus: despite the low confidence, there is definite value in seasonal rainfall forecasts Hagedorn, R. , and L. A. Smith, 2008: Communicating the value of probabilistic forecasts with weather roulette. Meteor. Appl. , DOI: 10. 1002/met. 92.

Increasing the value of climate forecasts How to from a technical perspective? What kind of climate forecasts best inform early-warning systems? Reliable = forecast probabilities correspond well with observed frequencies Timely Understandable language Salient = forecast must relate to an outcome of direct interest to the user Sharp = assigned probabilities are high enough for effective sectoral resources allocation Cost-effective and sustainable = requiring guaranteed continuity of operations in the face of frequent understaffing and limited funding Context

Increasing the value of climate forecasts Adding a reference = setting the climate context for upcoming events example: detailed climatologies 48

Increasing the value of climate forecasts Before the forecasts – what is the context? Context: Know where you are at now and where you’re supposed to be next Climate monitoring and climate impacts database: as one of the core functions of the Caribbean Regional Climate Centre is the monitoring of the Caribbean’s climate conditions. We operationally focus on seasonal scales to support the forecasts. A climate impacts database then allows tracking of on-the-ground conditions as well as research into forecasting climate impacts. Climatological norms for temperature and rainfall (and increasingly on derived information). RATIONALE: The perceptions of Caribbean climate and what drives climate are very vague. Therefore, there is not much appreciation for what usually happens in coming months.

Rainfall reference climatologies & monitoring Monthly and 3 -monthly climatologies For both reference period 1981 -2010 & for entire period of record 26 countries and territories

El Niño/La Niña – Drivers of climate variability Out of an El Niño… and possibly into a La Niña? For Guyana: Secondary wet season (November till January) likely wet to very wet.

Increasing the value of climate forecasts Continued Cari. COF support Understandable forecasts: Producing alert maps pertaining to impactful climate hazards. RATIONALE: Climate outlooks, though derived from probabilistic forecasts, may be expressed in a format co-developed by Met Services and partner stakeholders. This makes the forecast information more actionable. Communicate effectively and train user stakeholders through stakeholder engagement, dissemination using different media channels.

Increasing the value of climate forecasts Adding persistence = adding certainty to an uncertain forecast of slow onset events slow onset example: drought outlooks 53

Increasing the value of climate forecasts Adding persistence - example: drought forecasts HOW MUCH RAINFALL WE NORMALLY EXPECT 54

Increasing the value of climate forecasts Adding persistence - example: drought forecasts HOW MUCH RAINFALL WE EXPECTED IN JFM 55

Increasing the value of climate forecasts Adding persistence - example: drought forecasts TRADITIONAL SEASONAL RAINFALL OUTLOOK 56

Increasing the value of climate forecasts Adding persistence - example: drought forecasts HOW MUCH RAIN FELL SO FAR? 57

Increasing the value of climate forecasts Adding persistence - example: drought forecasts SAME SEASONAL RAINFALL FORECAST … but adding in observed rainfall surplus/deficit 58

Increasing the value of climate forecasts Adding persistence - example: drought forecasts THIS IS WHAT ACTUALLY HAPPENED!! 59

Hazard/Theme specific Products Cari. COF drought alerts Drought alert map

Increasing the value of climate forecasts Forecasting rainfall occurrence = adding certainty to an uncertain forecast of fast onset events fast onset example: wet days and wet spells outlooks 61

Increasing the value of climate forecasts Continued Cari. COF support Reduced uncertainty: Sub-seasonal information such as extreme rainfall and dry spell occurrence within a season. RATIONALE: Technically, predicting the rainfall occurrence can be done with much less uncertainty than the actual rainfall sum over a season. Plus, from the user perspective, a season with long dry spells and brief, extremely wet episodes will be more impactful to most sectors than a season with regular, light rainfall (except for ground water recharge). Thus, sub-seasonal information enhances the seasonal forecasts. Cari. COF Wet Days and Wet Spells Outlooks addresses this need, while forming a stepping stone to forecasting flash flood potemtial.

Forecasting wet day frequency within a season Wet day = day with precipitation of at least 1 mm Their frequency (or numbers) within a season indicate whether: a) rainfall well spread (high number) or concentrated (low number) if season as a whole is wet b) scant rainfall will be spread (low number) or concentrated (very low number) if season as a whole is dry Frequency shift decrease no change increase 63

Forecasting wet day frequency within a season FORECAST: Region-wide, June to August rainfall expected to be below- to normal, with fewer wet days than usual. Frequency shift decrease no change increase 64

2010 JJA Rainfall at Hewanorra, St. Lucia JUNE JULY AUGUST SEPT

2010 JJA Rainfall at Hewanorra, St. Lucia JUNE JULY AUGUST 7 – day wet spells (20% wettest) SEPT

NUMBER OF 7 -DAY WET SPELL 12 10 8 6 CLIMATOLOGY 4 2 Hewanorra Average number of 7 -day wet spells at Hewanorra

NUMBER OF 7 -DAY WET SPELL 12 10 8 6 predicted number of 7 -day wet spells at Hewanorra for JJA 2015 4 2 Hewanorra

NUMBER OF 7 -DAY WET SPELL 12 Average range of 7 -day wet spell events: events 3 to 7 10 8 Predicted range of 7 -day wet spell events for JJA 2015: events 6 4 2 Hewanorra 2 to 5

12 10 Climatological range of 7–day wet spell events during June, July, and August PREDICTED 7–day wet spell events during June , July, and August 2015 Top 20% wettest 7 day wet spells 8 6 4 2 Antigua St. Lucia Barbados Hewanorra CIMH GAIA Belize Cayman Dominica Jamaica Martinique Canefield. Worthy Park FDF

Forecasting shifts in wet spell frequency within a season 7 -day wet spell = period of 7 consecutive days with precipitation in the top 20% 71

Forecasting shifts in wet spell frequency within a season JJA 2015 frequency of 7 -day wet spells 7 -day wet spell = period of 7 consecutive days with precipitation in the top 20% 72

Forecasting shifts in very wet spell frequency within a season JJA 2015 frequency of 7 -day wet spells 7 -day wet spell = period of 7 consecutive days with precipitation in the top 20% JJA 2015 frequency of 7 -day very wet spells 7 -day very wet spell = period of 7 consecutive days with precipitation in the top 10% 73

Forecasting shifts in very wet spell frequency within a season JJA 2015 frequency of 7 -day wet spells FORECAST: Region-wide, June to August rainfall expected to be below- to normal, with fewer wet days and wet spells than usual. Reduced flood potential JJA 2015 frequency of 7 -day very wet spells throughout the region (with the exception of the Bahamas) 74

June to August 2015 Forecast Antigua (VC Bird) Aruba (Beatrix) Barbados (CIMH) Barbados (GAIA) Belize (C. Farm) Cayman Cuba (Punta Maisi) Dominica (Canefield) Dominica (Douglas Charles) Dom. Republic (Las Americas) Grenada (MBIA) Jamaica (Worthy Park) Martinique (FDF Desaix) Puerto Rico (San Juan) St. Lucia (Hewanorra) St. Maarten (TNCM) St. Vincent (ET Joshua) Suriname (Zanderij) Trinidad (Piarco) Tobago (ANR Robinson) Number of wet days Number of 7 -day wet spells (20% wettest) Number of 7 -day very wet spells (10% wettest) Climatology Forecast 26 -41 7 -17 34 -50 36 -48 34 -52 27 -41 7 -15 48 -66 53 -68 19 -34 42 -50 26 -38 48 -63 31 -48 23 -35 3 -11 27 -40 29 -43 26 -40 24 -39 5 -12 43 -60 49 -63 18 -30 40 -51 18 -31 46 -59 29 -48 1 -6 0 -5 2 -7 3 -8 2 -7 0 -3 3 -10 2 -6 1 -7 Na. N 2 -6 3 -7 2 -6 1 -4 0 -3 2 -5 1 -4 2 -7 2 -6 0 -3 2 -14 1 -5 0 -4 Na. N 0 -5 2 -5 0 -3 0 -2 0 -4 1 -5 1 -4 0 -2 1 -6 0 -3 0 -4 Na. N 1 -4 1 -4 0 -2 0 -3 1 -4 0 -3 0 -2 0 -9 0 -2 0 -3 Na. N 1 -3 0 -3 41 -58 35 -46 3 -7 2 -5 0 -4 0 -2 30 -42 55 -67 52 -63 53 -64 42 -52 23 -40 52 -65 51 -65 44 -57 31 -46 1 -6 3 -8 3 -7 5 -8 3 -7 0 -4 2 -7 2 -6 3 -8 2 -7 0 -3 1 -5 3 -5 1 -4 0 -3 1 -5 2 -5 1 -4 brown - decrease in numbers, dark blue – increase in numbers

h. edu. bb Extreme wet spells frequency shifts Forecast for: May to July 2016 MJJ 2016 frequency of extreme (top 1%) 3 -day wet spells USUALLY: Up to 1 or 2 extreme wet spells between May & July. FORECAST: MJJ rainfall is likely to be above- to normal in the north-western islands and the Guianas, but below- to normal in ABC Islands, Belize, Hispaniola and the US C’bean Territories. Fewer wet days (low confidence) are expected across the region, but with more wet and very spells in the Lesser Antilles (medium confidence) and increased chance of extreme wet spells, especially in the east (medium to high confidence). IMPLICATION: Serious flash flood potential developing.

Increasing the value of climate forecasts Communication = improving the uptake of climate outlook information for risk management by our users risk management examples: climate bulletins & stakeholder interaction 77

Where can we read about this? Our bulletins aim to offer more digestible overviews : Cari. COF Climate Outlook Newsletter Click here Caribbean Coral Reef Watch Click here Caribbean Drought Bulletin Click here Regional Agroclimatic Bulletin Click here

Drought alert outlook 79 Early drought detection Drought monitoring Capacity building and education Informing regional and national drought policy

Drought alert outlook 80 Drought situation – monitoring and prediction based on SPI Overview of main drought impacts Communication to stakeholders

Monthly Caribbean Drought Bulletin Announcement on drought + Overview of drought situation + rainfall intensity over past 1, 3, 6 and 12 months + Headline of impacts http: //rcc. cimh. edu. bb 81

Monthly Caribbean Drought Bulletin 6 -month Drought alert outlook (3 months in advance) + Update of the Hydrological Year / Wet Season Drought alert outlook (Jun to May / Dec to Nov) + References 82 http: //rcc. cimh. edu. bb

The Cari. COF - Innovative ways to communicate • Any form of message communication can be explored • Awareness and product interpretation built through interactive and other practical activity • Message communicated in drama, song and dance…

Increasing the value of climate forecasts Common standard = uniformity of climate outlooks generates an authoritative voice towards our users examples: Caribbean Outlook Generator (CAROGEN) 84

Increasing the value of climate forecasts Continued Cari. COF support Improving Cost-effectiveness and sustainability: IT automation of production (and to a limited extent mapping of) climate outlooks. RATIONALE: As is commonly the case in forecasting any aspect of nature or society, a larger number of well founded forecast experiments reduces uncertainty, but increases human resource needs. Automation largely cancels this strain on HR. CAROGEN addresses this need, while also allowing statistical insight into the countries’ climate

TIME FOR A BREAK

CIMH Early Warning Information Products: An Overview Dr. Cedric J. VAN MEERBEECK – Climatologist (cmeerbeeck@cimh. edu. bb) More climate information and products on http: //rcc. cimh. edu. bb Other major partners: Dominica In-Country Workshop: Mapping provider capacity and user needs for climate services Jimmit, March 18 th, 2016

Key Messages 1. Hot in 2016: uncomfortable at times, especially in August and September, with record high temperatures in many places. 2. Drought of 2015 -’ 16 is fading: was region-wide one of the worst droughts. 3. Why the heat and drought? A very strong El Niño – which faded out in May – largest cause for drought earlier AND added extra heat to the ocean and atmosphere. 4. Current outlook: • • • 5. further drought alleviation in coming months (wet season) for most. Hurricane season normal to active this year. Heat intense until end of October, especially during dry spells. Be prepared for: flash flood and land slides during the wet season.

The 2015 -’ 16 Drought

Recently in the news – drought!! 2015: record dry in many places, with drought continuing into 2016 • Antigua From August: “No surface water exploitable, 83% desalination” (Antigua & Barbuda Met Services) From October: 100% desalinated water • Barbados Potworks Dam in Antigua Mona reservoir in Jamaica record driest year resulting in water outages in central and northern part of the island. • Dominica Tropical Storm Erika amidst drought: 1. 3 billion EC$ damage and losses. • St. Kitts & Nevis Carraizo reservoir in Puerto Rico water rationing resumes in January 2016 after second driest year on record. No water sold to cruise ships since September. • St. Lucia two driest years on record in 2014 & 2015 led to water restrictions for all households. https: //anumetservice. wordpress. com/2015/08/17/antigua-is-out-of-surface-water-again/ http: //jamaica-gleaner. com/article/lead-stories/20150807/drought-affecting-lives-across-caribbean http: //www. washingtontimes. com/news/2015/aug/5/parched-caribbean-faces-widespread-drought-waters/? page=all http: //jamaica-gleaner. com/article/lead-stories/20150706/st-thomas-struggles-cope-water-crisis http: //www. jamaicaobserver. com/news/Drought-fuels-bush-fires-in-Manchester

Recently in the news – drought!! 2015: record dry in many places, with drought continuing into 2016 • Belize Millions $ losses in agriculture. • Cuba most severe drought since 2004 with water deficits in 45% of country and 100, 000 people depending on water delivery by tanks. Potworks Dam in Antigua Mona reservoir in Jamaica • Dominican Republic 11% decrease in agricultural production. • Guyana Water Inc. updated water rationing and management practices for Georgetown. • Haïti Carraizo reservoir in Puerto Rico 200, 000 families affected by ongoing drought, with a 30% reduction in harvest over 2015 • Puerto Rico streamflows well below average, eastern PR under drought, rationed water supply to San Juan up till October. (PR Met Service) https: //anumetservice. wordpress. com/2015/08/17/antigua-is-out-of-surface-water-again/ http: //jamaica-gleaner. com/article/lead-stories/20150807/drought-affecting-lives-across-caribbean http: //www. washingtontimes. com/news/2015/aug/5/parched-caribbean-faces-widespread-drought-waters/? page=all http: //jamaica-gleaner. com/article/lead-stories/20150706/st-thomas-struggles-cope-water-crisis http: //www. jamaicaobserver. com/news/Drought-fuels-bush-fires-in-Manchester

Recently in the news – in Suriname “C h ick en far m los ing sto ck t oh ea t” 2016: Dry earlier this year, hot during the long dry season http: //www. nieuws-suriname. nl/de-enorme-hitte-en-droogte-in-suriname/ http: //www. de-surinaamse-krant. com/2016/01/aanhoudende-droogte-in-suriname-begint. html http: //www. mo. be/wereldblog/aanhoudende-droogte-suriname-zorgwekkend

How do we know? Our drought monitoring products E. g. the map below shows 12 -month drought severity between May 2015 and April 2016 LEGEND Exceptionally Dry Normal Exceptionally Wet Source: Caribbean Drought Bulletin most severe usual most severe very rare common very rare

How do we know?

What’s the latest? Dec. 2015 - Aug. 2016 - Cultuurtuin (Suriname) Wet in April-May-June Climatological norm This year

What’s the latest? Drought alleviation for most affected areas Long-term drought severity by the end of April 2016 Long-term drought severity by the end of August 2016

What’s the latest? Hot in Aug. 2016 - warmer throughout the region

What’s the outlook?

Drought will gradually be alleviated Drought short to midterm drought IMPLICATIONS: long-term drought • Surface and soil wetness and river flow to increase where still low. • Large water reservoirs recharging. Drought alert & action levels

How wet or dry will the next three to six months be? October to December 2016 Drier Usual January to March 2017 Wetter Drier Usual FORECAST: 1) above- to normal in the ABC Islands, Belize, the Greater Antilles and W Guianas, below- to normal in E Guianas; 2) Roughly 80% confidence that second half of wet season will be wetter or usual E & S C’bean, but drier in extreme NW. IMPLICATIONS: • Moist conditions may favour mosquito breeding; Wetter • Soil drying up slower than in 2014 and 2015 dry seasons in E & S C’beam.

How often will it rain in the next three months? October to December 2016 OND 2016 frequency of wet days outlook USUALLY: Across the region during October. November-December, about 30 to 50 days are wet days. FORECAST: up to 20% more wet days (medium to high confid. ) in the territories north of 16°N. IMPLICATION: • Disruptions of outdoor activities by rainfall occurring regularly, but fewer than usual in Aruba, Grenada, parts of Guyana and Trinidad & Tobago. caricof@cimh. edu. bb

How many wet spells will we see? October to December 2016 OND 2016 frequency of 7 -day wet spells wet and very wet spells outlook USUALLY: Near the end of the wet season (Guianas: transition from dry to wet season; ABC Islands: wet season), roughly 3 to 5 wet spells (ABC Islands: 3 to 9) occur from October to December, of which 1 to 4 are very wet (ABC Islands: 1 to 7). FORECAST: More wet spells in most places (medium confidence). IMPLICATION: Effective recharging of large water reservoirs during wet season months. caricof@cimh. edu. bb

Extreme wet spells frequency shifts Forecast for: October to December 2016 OND 2016 frequency of extreme (top 1%) 3 -day wet spells USUALLY: Up to 1 or 2 extreme wet spells occur from October to December (up to 3 in ABC Islands). FORECAST: no large shift in numbers of extreme wet spells is forecast throughout the Caribbean, except for an increase in Jamaica and Suriname (medium confid. ). IMPLICATION: Some flash flood potential developing, especially in northern Guianas. caricof@cimh. edu. bb

How hot will the next three to six months be? Jan-Feb-Mar 2017 Oct-Nov-Dec 2016 Night time Mild Usual Day time Hot Mild Usual FORECAST: 1) temperatures between October and December will be above- to normal; both nights and days will be warm. 2) Less certain January to March, but still likely above- to normal. Hot IMPLICATIONS: • Heat intense until end of October in S C’bean. • Enhanced health risk from heat exposure. • Higher than usual energy costs for cooling.

The Climate Outlook in Summary • For October-November-December 2016: – – – Initially uncomfortably hot in Southern Caribbean, especially in case of dry spells Surface water reservoirs recharge, soil moisture replenishment and increased river flow. Water shortage related problems in agriculture to disappear in those few places still affected. Extremely wet spells may occur. Serious potential for flash flooding and landslides Hurricane activity: more active than past 5 years, expected to slowly wind down during October and November. • For January-February-March 2016: – Possibly wetter than usual in E & S C’bean, making dry season in the islands less intense there; but the opposite in the far NW C’bean. – Excessive rainfall could increase risk of flash flooding & long-term flooding in Guianas.

Where can we read about this? Our bulletins aim to offer more digestible overviews: Cari. COF Climate Outlook Newsletter Click here Caribbean Coral Reef Watch Click here Caribbean Drought Bulletin Click here Regional Agroclimatic Bulletin Click here

Thank you rcc. cimh. edu. bb