Cari COF Climate Outlook NovemberDecember 2014January 2015 and

Cari. COF Climate Outlook November-December 2014/January 2015 and February-March-April 2015 Coordination – CIMH – Dr. Cédric J. Van Meerbeeck Participating territories Antigua & Barbuda, Aruba, Bahamas, Barbados, Belize, Cayman Islands, Cuba, Curaçao, Dominican Republic, French Guiana, Grenada, Guadeloupe, Guyana, Jamaica, Martinique, Puerto Rico, St. Barth’s, St. Lucia, St. Maarten/St. Martin, St. Vincent & the Grenadines, Suriname, Trinidad & Tobago and the US Virgin Islands

Regional Climate Outlook Forums Cari. COF - FCAC Cari. COF FCAC XLII FCAC – Ciudad de Guatemala, 09/10 -04 -2014 2

RAINFALL

NDJ rainfall

CPT probabilistic NDJ rainfall forecast CCA experiments: 1) Predictor is observed SST over the tropical Atlantic and Pacific over September (data source: NOAA ERSSTv 3 b, obtained from the IRI data library) 2) Predictor is observed SST over the tropical North Atlantic over September 3) Predictor is predicted SST over the tropical Atlantic and Pacific over NDJ (data source: NOAA CPC CFSv 2 , August 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) 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.

Experiment 1

CPT rainfall forecast Data: Predictor = August SST observations 1952 -2014 140 E-20 W and 30 N-20 S (i. e. Tropical Pacific & Atlantic) [Source: ERSSTv 3 b, from IRI data library] Predictand = OND rainfall for 156 Caribbean stations 1971 -2013 Parameters: CCA modes: max = 5; optimum = 1 X modes: max = 8; optimum = 1 Y modes: max = 8; optimum = 1 Missing Y data threshold: 20% missing values, 10% missing stations, leading to 123 stations included in hindcast and forecast Goodness index (transformed Y) = !!0. 082 Very Limited!!

ROC area maps

CCA modes

Experiment 1 NDJ Rainfall Forecast Tercile maps Normal Below Above

Experiment 2

CPT rainfall forecast Data: Predictor = September SST observations 1979 -2014 90 -20 W and 0 -30 N (i. e. tropical N Atl. ) [Source: ERSSTv 3 b, from IRI data library] Predictand = NDJ rainfall for 156 Caribbean stations 1979 -2013 Parameters: CCA modes: max = 5; optimum = 5 X modes: max = 8; optimum = 6 Y modes: max = 8; optimum = 6 Missing Y data threshold: 20% missing values, 10% missing stations, leading to 143 stations included in hindcast and forecast Goodness index (transformed Y) = !0. 118 Limited!

ROC area maps

CCA modes

Experiment 2 NDJ Rainfall Forecast Tercile maps Normal Below Above

Experiment 3

CPT rainfall forecast Data: Predictor = NDJ simulated SST 1982 -2014 140 E-20 W and 30 N-20 S (i. e. tropical Pacific & Atlantic) [Source: NOAA CPC – CFSv 2, September initialisatio Predictand = NDJ rainfall for 156 Caribbean stations 1983 -2013 Parameters: CCA modes: max = 5; optimum = 1 X modes: max = 8; optimum = 7 Y modes: max = 8; optimum = 8 Missing Y data threshold: 20% missing values, 10% missing stations, leading to 143 stations included in hindcast and forecast Goodness index (transformed Y) = !0. 100 limited!

ROC area maps

CCA modes

Experiment 3 NDJ Rainfall Forecast Tercile maps Above Normal Below

Experiment 4

Data: Predictor = September CPT rainfall forecast OND simulated SST 90 -20 W and 0 -30 N (i. e. tropical N Atl. ) [Source: NOAA CPC – CFSv 2, initialisation] Predictand = OND rainfall for 156 Caribbean stations 1983 -2013 Parameters: CCA modes: max = 5; optimum = 1 X modes: max = 8; optimum = 8 Y modes: max = 8; optimum = 4 Missing Y data threshold: 20% missing values, 10% missing stations, leading to 143 stations included in hindcast and forecast Goodness index (transformed Y) = !0. 152 Limited!

ROC area maps

CCA modes

Experiment 4 NDJ Rainfall Forecast Tercile maps Normal Below Above

Experiment 5

Data: Predictor = CPT rainfall forecast NDJ simulated rainfall 100 -40 W and 5 S-35 N (i. e. broader Caribbean) [Source: ECHAM 4. 5 ensemble 24 – from IRI data library] Predictand = NDJ rainfall for 156 Caribbean stations 1979 -2013 Parameters: CCA modes: max = 5; optimum = 1 X modes: max = 5; optimum = 1 Y modes: max = 5; optimum = 1 Missing Y data threshold: 20% missing values, 10% missing stations, leading to 143 stations included in hindcast and forecast Goodness index (transformed Y) = !0. 129 Limited!

ROC area maps

CCA modes

Experiment 5 NDJ Rainfall Forecast Tercile maps Normal Below Above

FMA rainfall

CPT probabilistic JFM rainfall forecast CCA experiments: 1) Predictor is observed SST over the tropical Atlantic and Pacific over Sep (data source: NOAA ERSSTv 3 b, obtained from the. IRI data library) 2) Predictor is predicted SST over the tropical Atlantic and Pacific over FMA (data source: NOAA CPCFSv 2, Sep(initialisation)

Experiment 1

CPT FMA rainfall forecast Data: Predictor = Sep SST observations 1979 -2013 140 E-20 W and 30 N-20 S (i. e. tropical Pacific & Atlantic) [Source: ERSSTv 3 b, from IRI data library] Predictand = FMA rainfall for 165 Caribbean stations 1979 -2013 Parameters: CCA modes: max = 5; optimum = 3 X modes: max = 8; optimum = 6 Y modes: max = 8; optimum = 7 Missing Y data threshold: 20% missing values, 10% missing stations, leading to 133 stations included in hindcast and forecast Goodness index (transformed Y) = 0. 126 ! Limited !

ROC area maps

CCA modes

Experiment 1 FMA Rainfall Forecast Tercile maps NORMAL BELOW ABOVE

Experiment 2

CPT FMA rainfall forecast Data: Predictor = FMA simulated SST 1983 -2014 140 E-20 W and 30 N-20 S (i. e. tropical Pacific & Atlantic) [Source: NOAA CPC – CFSv 2, sep initialisation] Predictand = FMA rainfall for 165 Caribbean stations 1979 -2013 Parameters: CCA modes: max = 5; optimum = 5 X modes: max = 8; optimum = 8 Y modes: max = 8; optimum = 7 Missing Y data threshold: 20% missing values, 10% missing stations, leading to 153 stations included in hindcast and forecast Goodness index (transformed Y) = 0. 181 ! Limited!

ROC area maps

CCA modes

Experiment 2 FMA Rainfall Forecast Tercile maps NORMAL BELOW ABOVE

Supporting probabilistic rainfall forecasts CARICOF national/sub-regional expert views: 1) Puerto Rico & USVI: NDJ FMA A = 56% ; N = 28% ; B = 16% A = 50% ; N = 30% ; B = 20% Note: CPT CCA showed limited skill. Under El Niño conditions, a wetter forecast could be expected across PR and USVI during the dry season. 2) Guadeloupe: NDJ Martinique : NDJ French Guiana: NDJ French Northern Antilles: NDJ A = 35% ; A = 32% ; A = 30% ; A = 35% ; N = 35% ; N = 36% ; N = 38% ; N = 35% ; B = 30% B = 32% B = 30% Note: For the Lesser Antilles there is an agreement between several dynamic models (Eurosip, JMA) ) for a drier than normal forecast in NDJ 2014 -2015. But the statistic SST/RR models (with low skill on the leeward islands) show above terciles. (predictor stay SST over tropical North Atlantic and Pacific, with a low but positive ENSO anomaly). Over French Guiana : The statistic model has good skill and says above. The dynamic models exhibit a drier than normal forecast. Over JAS, the drought situation was: Normal on St. Barth’s and very dry in St. Martin, normal to dry in Guadeloupe, normal (windward side) to extremely dry (leeward side) in Martinique. Wet to exceptionally conditions in French Guiana. 3) Barbados: NDJ FMA A = 34%; N = 34% ; B = 32% A = %; N = % ; B = % Note: CPT CCA with ERRSTs and CFSv 2 SSTs showed limited skill and poor discrimination with ROCs around 0. 5. ECMWF Prognostic map is indicating normal rainfall for the Eastern Caribbean. A slight indicator for above normal rainfall is noted over the northern half of the island chain. 4) St. Maarten: NDJ A = 33%; N = 35% ; B = 32% Note: Global models are not all in agreement this time around similarly, my experiment do not all give the same indication. Goodness Index was poor except in the case of predicted ssts.

Supporting probabilistic rainfall forecasts CARICOF national/sub-regional expert views: 5) Trinidad and Tobago: NDJ FMA A = 36% ; N = 34% ; B = 30% A = 33% ; N = 36% ; B = 31% Note: Seasonal climate probabilistic models (IRI, EUROSIP-MM, NMME, ECMWF, UKMET, A) show some divergence in their forecasts for Trinidad and Tobago with IRI indicating Below Normal for NDJ; NMME, EUROSIP, UKMET and ECMWF indicating equal chances for BN, AN, N; while individual NMME models CFSV 2, and GFDL favour Above Normal rainfall for NDJ near Trinidad and Tobago. [missing text] as well as Australian rainfall patterns, continue to show some El Niño-like signatures, but remain in the neutral range. ENSO remains in neutral status but the NINO 3. 4 region continues to be warmer than average with some connectivity to our region through climate signals. Models suggest there still a reasonable chance for light to moderate El Nino conditions by the end of NDJ. Even though El Nino thresholds have not been met, Trinidad and Tobago rainfall patterns within recent seasons have shown El Nino-like signatures, being drier and warmer than average. SSTs in and around Trinidad and Tobago have warmed slightly but remain cooler than average for this time of the year. SSTs just east of T&T in the North Atlantic Ocean are forecast to be above average during NDJ while those further east are expected to cooler than average. The divergence among individual GPC models and the recent and forecast climatic environment reduces confidence in the NDJ _CPT results. 6) Jamaica: NDJ A = 34% ; N = 31% ; B = 35% Note: Through the period, the forecasts from the Global Dynamic Models indicate below normal rainfall with warmer than normal temperatures likely to continue across the Caribbean. However, findings from the CPT showed mixed results due to the very low confidence in the sea surface temperature signals. Fifteen rainfall stations were examined across the island, of which eight are likely to receive near normal to below normal rainfall, while seven stations are likely to have above normal rainfall during the period. Stations across southern parishes are indicating low forecast confidence of above normal rainfall. Pending El Niño conditions as well as drier than normal atmospheric conditions across much of the Caribbean continues to be the main reasons forecasting below average rainfall as well as warmer than normal temperature across most areas of the island.

Supporting probabilistic rainfall forecasts CARICOF national/sub-regional expert views: 7) Suriname: Note: 8) NDJ FMA CPT CCA showed very limited skill (Pearson’s corr 0. 35). IRI predicts 40% BN for NDJ. Belize coast (except south) Belize (south/inland) Note: NDJ A = 40% ; N = 35% ; B = 25% A = 35% ; N = 40% ; B = 25% CPT CCA with limited skill using September ERSSTs. Forecast based on ecperiment and global models. 9)St. Vincent: Note: A = 62% ; N = 22% ; B = 16% A=%; N=%; B=% NDJ A = 35% ; N = 34% ; B = 32% CPT CCA with limited skill using ERSSTs and CFSv 2 SSTs. 10)Curaçao: ASO A=%; N=%; B=% NDJ FMA A = 27% ; N = 32% ; B = 41% A = 40% ; N = 36% ; B = 24% Note: 11) Grenada: Note: CPT CCA with moderate to moderate limited skill for NDJ.

Supporting probabilistic rainfall forecasts CARICOF national/sub-regional expert views: 12)Aruba: NDJ FMA A = 30% ; N = 43% ; B = 27% A=%; N=%; B=% Note: CPT CCA showed very limited skill for NDJ. Technical difficulties experienced in FMA experiments. Supporting info: ENSO forecast: The SINTEX-F model predictions indicate that the present weak El Niño will continue until early next year. The SST anomalies extended all the way along the equator will then gradually turn to an El Niño Modoki. (probability for N is suspicious, given that a correlation skill of >0. 9 would be required to obtain a value of 43%) 13)St. Lucia: NDJ FMA A = 25% ; N = 30% ; B = 45% A = 25% ; N = 35% ; B = 40% Note: CPT CCA showed negative skill for NDJ, fair for FMA using ERSSTs. 14)Cayman: NDJ A = 44% ; N = 19% ; B = 37% Note: CPT CCA showed very limited to limited skill using Aug ERSSTv 3 b in terms of correlation skill, but good discrimination of B and A (ROC>0. 8) 15)Dominica: NDJ JFM A = 36% ; N = 32% ; B = 32% A=%; N=%; B=% Note: CPT CCA with limited skill using ERSST for September and moderate skill with NDJ CFSv 2’. 16)Guyana: NDJ A = 36% ; N = 35% ; B = 29% Note: CPT CCA with moderate skill using ERSST and CFSv 2 SST and moderate to good discrimination.

Cari. COF CPT output (background) and National/sub-regional rainfall outlooks

Cari. COF precipitation outlooks

Supporting probabilistic precipitation forecasts Eight data sources (incl. GPCs): 1) IRI multi-model probability forecast (only multi-model forecast that recalibrates & weights the models based on past performance); 2) UK Met Office GCM (UKMO) probability forecast; 3) European Center for Mid-range Weather Forecast GCM (ECMWF) and EUROSIP (multi-model) probability forecasts; 4) APEC Climate Center (APCC) multi=model probability forecasts; 5) WMO Lead Centre for LRF – MME. 6) CFSv 2 model 7) Météo. France Arpège model. 8) JMA model Rainfall outlooks are generated next to T 2 m outlooks (see Appendix). Rainfall tendencies from supporting models and local expertise: 1) 2) 3) ENSO conditions have recently been (warm-)neutral (SST anom. 0 -0. 5°C). Most new model runs suggest a progression into an El Niño by NDJ 2014, with an estimated overall confidence of 67% for NDJ and 58 -66% for FMA. Similarly, SSTs are slightly above average around N and E of the Caribbean, as well as further out eastward in the N Atlantic. However, this marks a small reversal in anomalies compared to previous months and may be (partly) due to weaker trade winds. associated with an unusual northward excursion of the ITCZ in late September and the first half of October as well as two periods of negative NAO. Into FMA, indications are that eastern portions of the Tropical North Atlantic could return to average SSTs. Thanks to slightly warmer SSTs around the Atlantic during NDJ, average to above-average air moisture may be the prevalent pattern, marking a shift from a pattern of drier air observed during much of the year. Further, the suggested weak El Niño may sustain increased vertical wind shear over the tropical North Atlantic as well as over eastern and southern Caribbean (including the Guianas). The signals of added moisture, but possibly stronger wind shear will counteract each other in producing anomalous rainfall. Testimony to the uncertainty as to which factor will dominate is the disagreement between some global models as to NDJ rainfall, with some suggesting below- to normal over the E Caribbean, and others rather above-normal.

Probabilistic NDJ rainfall forecast map D D A R T F

Probabilistic FMA rainfall forecast map D D A R T F

2 m TEMPERATURE

NDJ

CPT probabilistic NDJ 2 m temp. forecast CCA experiments: 1) Predictor is observed SST over the tropical Atlantic and Pacific over September (data source: NOAA ERSSTv 3 b, obtained from the IRI data library) 2) Predictor is observed SST over the tropical North Atlantic over September 3) Predictor is predicted SST over the tropical Atlantic and Pacific over NDJ (data source: NOAA CPC CFSv 2 , October initialisation) 4) Predictor is predicted SST over the tropical North Atlantic over NDJ 5) Predictor is predicted T 2 m over the Caribbean over NDJ (data source: ECHAM 4. 5 ensemble 24, obtained from the IRI data library)

Experiment 1

Data: Predictor = CPT T 2 m forecast September SST observations 1979 -2014 140 E-20 W and 30 N-20 S (i. e. Tropical Pacific & Atlantic) [Source: ERSSTv 3 b, from IRI data library] Predictand = NDJ T 2 m for 51 Caribbean stations 1979 -2013 Parameters: CCA modes: max = 5; optimum = 2 X modes: max = 8; optimum = 2 Y modes: max = 8; optimum = 3 Missing Y data threshold: 20% missing values, 10% missing stations, leading to 36 stations included in hindcast and forecast Goodness index (transformed Y) = 0. 237 Moderate

ROC area maps

CCA modes

Experiment 1 NDJ T 2 m Forecast Tercile maps ABOVE NORMAL BELOW

Experiment 2

Data: Predictor = CPT T 2 m forecast September SST observations 1979 -2014 90 -20 W and 030 N (i. e. tropical N Atl. ) [Source: ERSSTv 3 b, from IRI data library] Predictand = NDJ T 2 m for 51 Caribbean stations 1979 -2013 Parameters: CCA modes: max = 5; optimum = 3 X modes: max = 8; optimum = 3 Y modes: max = 8; optimum = 8 Missing Y data threshold: 20% missing values, 10% missing stations, leading to 36 stations included in hindcast and forecast Goodness index (transformed Y) = 0. 206 Moderate

ROC area maps

CCA modes

Experiment 2 NDJ T 2 m Forecast Tercile maps ABOVE NORMAL BELOW

Experiment 3

Data: Predictor = CPT T 2 m forecast NDJ simulated SST 1982 -2014 140 E-20 W and 30 N-20 S (i. e. tropical Pacific & Atlantic) [Source: NOAA CPC – CFSv 2, Oct initialisation] Predictand = NDJ T 2 m for 51 Caribbean stations 1982 -2013 Parameters: CCA modes: max = 5; optimum = 2 X modes: max = 8; optimum = 2 Y modes: max = 8; optimum = 3 Missing Y data threshold: 20% missing values, 10% missing stations, leading to 35 stations included in hindcast and forecast Goodness index (transformed Y) = 0. 194 !Limited!

ROC area maps

CCA modes

Experiment 3 NDJ T 2 m Forecast Tercile maps ABOVE NORMAL BELOW

Experiment 4

Data: Predictor = CPT T 2 m forecast NDJ simulated SST 90 -20 W and 0 -30 N (i. e. tropical N Atl. ) [Source: NOAA CPC – CFSv 2, Oct initialisation] Predictand = NDJ T 2 m for 51 Caribbean stations 1982 -2013 Parameters: CCA modes: max = 5; optimum = 4 X modes: max = 8; optimum = 6 Y modes: max = 8; optimum = 7 Missing Y data threshold: 20% missing values, 10% missing stations, leading to 35 stations included in hindcast and forecast Goodness index (transformed Y) = 0. 228 Moderate

ROC area maps

CCA modes

Experiment 4 NDJ T 2 m Forecast Tercile maps ABOVE NORMAL BELOW

Experiment 5

Data: Predictor = CPT T 2 m forecast NDJ simulated T 2 m 100 -40 W and 5 S-35 N (i. e. broader Caribbean) [Source: ECHAM 4. 5 ensemble 24 – from IRI data library] Predictand = NDJ T 2 m for 51 Caribbean stations 1979 -2013 Parameters: CCA modes: max = 3; optimum = 2 X modes: max = 5; optimum = 5 Y modes: max = 5; optimum = 2 Missing Y data threshold: 20% missing values, 10% missing stations, leading to 36 stations included in hindcast and forecast Goodness index (transformed Y) = 0. 149 !Limited!

ROC area maps

CCA modes

Experiment 5 NDJ T 2 m Forecast Tercile maps ABOVE NORMAL BELOW

FMA T 2 m

CPT probabilistic FMA 2 m temp. forecast CCA experiments: 1) Predictor is observed SST over the tropical Atlantic and Pacific over SEP (data source: NOAA ERSSTv 3 b, obtained from the. IRI data library) 2) Predictor is predicted SST over the tropical Atlantic and Pacific over FMA (data source: NOAA CPCFSv 2, October(initialisation)

Experiment 1

Data: Predictor = CPT FMA T 2 m forecast SEPTEMBER SST observations 1979 -2013 140 E-20 W and 30 N-20 S (i. e. tropical Pacific & Atlantic) [Source: ERSSTv 3 b, from IRI data library] Predictand = FMA T 2 m for 51 Caribbean stations 1979 -2014 Parameters: CCA modes: max = 5; optimum = 2 X modes: max = 8; optimum = 2 Y modes: max = 8; optimum = 3 Missing Y data threshold: 20% missing values, 10% missing stations, leading to 36 stations included in hindcast and forecast Goodness index (transformed Y) = 0. 327 Good

ROC area maps

CCA modes

Experiment 1 FMA T 2 m Forecast Tercile maps ABOVE NORMAL BELOW

Experiment 2

Data: Predictor = CPT FMA T 2 m forecast FMA simulated SST 1983 -2014 140 E-20 W and 30 N-20 S (i. e. tropical Pacific & Atlantic) [Source: NOAA CPC – CFSv 2, Oct initialisation] Predictand = FMA T 2 m for 51 Caribbean stations 1979 -2013 Parameters: CCA modes: max = 5; optimum = 2 X modes: max = 8; optimum = 5 Y modes: max = 8; optimum = 5 Missing Y data threshold: 20% missing values, 10% missing stations, leading to 37 stations included in hindcast and forecast Goodness index (transformed Y) = 0. 364 Good

ROC area maps

CCA modes

Experiment 2 FMA T 2 m Forecast Tercile maps ABOVE NORMAL BELOW

Supporting probabilistic T 2 m forecasts CARICOF national/sub-regional expert views: 1) Note: 2) Note: 3) Note: 4) Note: Cayman: NDJ A = 74% ; N = 16% ; B = 9% CPT CCA with ERSSTs showed good skill (GI 0. 451) and discrimination (ROCs ~0. 85). St. Lucia: NDJ FMA A = 50% ; N = 30% ; B = 20% A = 85% ; N = 10% ; B = 5% CPT CCA showed fair to good skill. Puerto Rico & USVI: NDJ FMA A = 55% ; N = 25% ; B = 20% A = 65% ; N = 25% ; B = 10% CPT CCA with ERSSTv 3 b for August showed limited skill. St. Maarten: NDJ A = 75%; N = 20% ; B = 5% CPT CCA with ERSSTv 3 b for Sep showed fair skill (GI 0. 36). Temperature will continue to be above normal. 5) Trinidad and Tobago: NDJ A = 56% ; N = 33% ; B = 11% (Tmax) A = 58% ; N = 29% ; B = 13% (Tmin) 6) Jamaica: NDJ A = 65% ; N = 25% ; B = 10% 7) Belize: NDJ A=%; N=%; B=% NDJ A = 26% ; N = 46% ; B = 29% Note: 8) Note: 9) Note: based on global models. St. Vincent: CPT CCA with fair skill using ERSSTs and good skill using CFSv 2 NDJ SSTs. Dominica: NDJ A = 48% ; N = 19% ; B = 34% CPT CCA with very limited skill using ERSSTs. (climatology period 1996 -2010)

Supporting probabilistic T 2 m forecasts CARICOF national/sub-regional expert views: 1) Barbados: NDJ DJF A = 60% ; N = 28% ; B = 12% A=%; N=%; B=% Note: CPT CCA with Sep ERSSTs showed good skill and discrimination ROCs ~0. 8. Climatology 1979 -2014. 2) Guyana: NDJ A = 37% ; N = 31% ; B = 32% Note: CPT CCA with Sep ERSSTs limited good skill but moderate to good discrimination. 3) Curaçao: SON A=%; N=%; B=% Grenada: NDJ FMA A = 43% ; N = 16% ; B = 41% A = 28% ; N = 15% ; B = 56% Note: 4) Note: 5) CPT CCA with negative skill for NDJ and fair skill for FMA. Suriname: NDJ FMA A = 34% ; N = 29% ; B = 37% A=%; N=%; B=% Note: CPT CCA with limited skill. IRI suggests 80% for AN for NDJ (but with longer climatology). 6) Aruba: NDJ FMA A = 57% ; N = 35% ; B = 7% A=%; N=%; B=% Note: CPT CCA with fair skill and good discrimination ROCs close to 0. 8

Cari. COF CPT output (background) and National/sub-regional T 2 m outlooks

Probabilistic NDJ 2 m temperature forecast map D D A R T F

Probabilistic FMA 2 m temperature forecast map D D A R T F

Drought outlook

Supporting probabilistic SPI forecasts CARICOF national/sub-regional expert views: 1) Barbados: Note: ASONDJ (SPI < -1. 3) P = 4% Rel. Odds = ASONDJ (SPI < -1. 3) P = 44% Rel. Odds = ~8 ASONDJ (SPI < -1. 3) P = 16% Rel. Odds = ~2. 0 CPT CCA with very limited skill but good discrimination ROC 0. 85. . 6) Cayman: Note: P = 13% Rel. Odds = 1. 2 CPT CCA with moderate skill including two stations from Dominica and 1 from Guadeloupe (Le Raizet). 5) St. Lucia: Note: ASONDJ (SPI < -1. 3) CPT CCA with limited skill (GI 0. 185) and moderate discrimination ROC 0. 67. 4) Dominica: Note: Rel. Odds = CPT CCA with moderate skill (GI 0. 0. 224). 3) Grenada: Note: P = 6% CPT CCA with poor discrimination ROC 0. 56. 2) St. Vincent: Note: NDJ (SPI < -1. 3) ASONDJ (SPI < -1. 3) P = 16% Rel. Odds = ~2 CPT CCA with limited skill but excellent discrimination ROC 0. 97.

SPI outlook Aug to Jan – areas under immediate drought concern? Previous month’s update: drought evolving over Belize, Dominica and Martinique; drought possible in most other areas (left inset). This month’s update: drought evolving over SE Belize and Cayman; drought possible in most areas (below). Drought evolving over south-east Belize and Cayman 101

SPI outlook Aug 2014 to Jan 2015 – shorter- / longer-term concern? • Current drought situation: – Haïti, the leeward side of Martinique and St. Martin are in drought and have suffered water shortages. – Recent drought conditions in Belize, Dominica and Jamaica improved. • Shorter-term: – We expect the drought situation to possibly appear some of the islands and improve over others. • Longer-term: – We expect an El Niño to happen by the end of this year. El Niño often results in a drier rainy and dry season. – Water shortages could occur in the dry season, which is our tourist season. CONSERVE WATER!! Especially in Belize, Cayman, Haïti, Jamaica, and E Caribbean. 102

Interpreting the drought alert levels Colour Alert level Meaning Suggested action level GREEN No Concern No drought concern Business as usual YELLOW Drought Watch Drought possible Keep updated, conserve water ORANGE Drought Warning Drought evolving Be prepared, conserve water. protect RED Drought Alert Drought of immediate concern Take action, ration water, protect 103

Appendix

US Climate Prediction center – Climate Diagnostics bulletin http: //www. cpc. ncep. noaa. gov/products/CDB/

NOAA CPC NAO index monitoring/forecasting http: //www. cpc. ncep. noaa. gov/products/precip/CWlink/pna/nao_index_ensm. shtml

Explanatory variables – IRI Predicted SSTs El Niño region Caribbean & tropical Atlantic http: //portal. iri. columbia. edu/portal/server. pt? open=512&obj. ID=585&Page. ID=0&cached=true&mode=2&user. ID=2

Explanatory variables – ECMWF Predicted Tropical SSTs http: //www. ecmwf. int/products/forecasts/d/charts/seasonal/forecast/seasonal_range_forecast/group_public/seasonal_charts_public_s st!sea%20 surface%20 temperature!1%20 month!Tropics!201401!ensemble%20 mean!/

Explanatory variables – CPC/IRI ENSO Forecast http: //iri. columbia. edu/our-expertise/climate/forecasts/enso/2013 -december-quick-look/

http: //iri. columbia. edu/our-expertise/climate/forecasts/enso/2013 -december-quick-look/

IRI – multi-model probabilistic rainfall forecast http: //iri. columbia. edu/our-expertise/climate/forecasts/seasonal-climate-forecasts/

EUROSIP – multi-model probabilistic rainfall forecast http: //www. ecmwf. int/products/forecasts/d/charts/seasonal/forecast/eurosip/group_euro_public/seasonal_charts_rain!tercile%20 s ummary!1%20 month!Tropics!201310!/

ECMWF – ensemble probabilistic rainfall forecast http: //www. ecmwf. int/en/forecasts/charts/seasonal/2 m-temperature-public-charts-long-range-forecast

UK Met Office – probabilistic rainfall forecast ABOVE NORMAL BELOW http: //www. metoffice. gov. uk/research/climate/seasonal-to-decadal/gpc-outlooks/glob-seas-prob

NOAA CPC – probabilistic rainfall forecast http: //www. cpc. ncep. noaa. gov/products/international/cfsv 2/camerica_prec_prob 2. shtml

APCC – multi-model probabilistic rainfall forecast http: //www. apcc 21. org/eng/service/6 mon/ps/japcc 030703. jsp

WMO LC for LRF-MME – Multi-model probabilistic rainfall forecast https: //www. wmolc. org/modules/data/plot_PMME. php? tm_id=1&cdepth=3&upnum=6&ca_id=101&s 1=3&s 2=1&t 1

WMO LC for LRF-MME – Individual deterministic rainfall forecast

Météo France Arpège – probabilistic rainfall forecast *** courtesy of Christophe Montout

Japan Met Agency – probabilistic rainfall forecast http: //ds. data. jma. go. jp/tcc/products/model/probfcst/4 m. E/fcst_gl. php

TRMM Monthly Rainfall Climatology January, February & March

TRMM Monthly Rainfall Climatology April, May & June

TRMM Monthly Rainfall Climatology July, August & September

TRMM Monthly Rainfall Climatology October, November & December

IRI – multi-model probabilistic T 2 m forecast http: //iri. columbia. edu/our-expertise/climate/forecasts/seasonal-climate-forecasts/

EUROSIP – multi-model probabilistic T 2 m forecast http: //www. ecmwf. int/products/forecasts/d/charts/seasonal/forecast/eurosip/group_euro_public/seasonal_charts_rain!tercile%20 s ummary!1%20 month!Tropics!201309!/

ECMWF – ensemble probabilistic T 2 m forecast http: //www. ecmwf. int/en/forecasts/charts/seasonal/2 m-temperature-public-charts-long-range-forecast

UK Met Office – probabilistic T 2 m forecast ABOVE NORMAL BELOW http: //www. metoffice. gov. uk/research/climate/seasonal-to-decadal/gpc-outlooks/glob-seas-prob

APCC – multi-model probabilistic T 2 m forecast http: //www. apcc 21. org/eng/service/6 mon/ps/japcc 030703. jsp

WMO LC for LRF-MME – Multi-model probabilistic T 2 m forecast https: //www. wmolc. org/modules/data/plot_PMME. php? tm_id=1&cdepth=3&upnum=6&ca_id=101&s 1=3&s 2=1&t 1

WMO LC for LRF-MME – Individual deterministic T 2 m forecast https: //www. wmolc. org/modules/data/plot_PMME. php? sm_id=1&tm_id=1&cdepth=3&upnum=6&ca_id=101&s 1=3&s 2=1&t 1=4

WMO Lead Centre for Long-Range-Forecast – individual GPC models inter-model consistency https: //www. wmolc. org/modules/data/plot. php? tm_id=1&cdepth=3&upnum=6&ca_id=83&s 1=3&s 2=3&t 1=4#