A sea of trouble

Author: Dimitris Herrera
I am a graduate student from a farming family in the Dominican Republic, and I study drought in the Caribbean. Having been home recently, I have taken some time to reflect on what the droughts and hurricanes over the last few years have meant for my family and friends back home.

Figure 1: A corn cropped small farm near San Juan, southwestern Dominican Republic. The crop dries out due to the lack of water for irrigation and no rain in the summer 2015. Although many farms in the Dominican Republic have access to flood irrigation, many farmers rely on the rainy season to crop their farms, making them highly vulnerable to both unexpected drought and extreme precipitation.

As most people interested in climate science will know, the 2017 Atlantic hurricane season has made history in the Caribbean. It has been recognized as the worst since 2005, in terms of material damage and death toll (NOAA, 2017); and it is indeed the most expensive hurricane season in the U.S., surpassing the devastating 2005 hurricane season. The most powerful and deadliest hurricanes of 2017, Maria and Irma, hit many countries across the Caribbean, causing significant casualties and property loss. Maria, for example, is considered as “the worst natural disaster in Dominica and Puerto Rico” by the governments of both islands, due to the extensive damage caused by this hurricane. To have an idea of the magnitude of Maria, up to 37.9 inches (962.7 mm) were accumulated as a result of this hurricane in certain areas of Puerto Rico. This amount of rain is about 57% of the annual mean precipitation for the island (~1687 mm (Daly et al, 2003)).

Regardless of the effects of Irma and Maria in the Caribbean, less than a year before of these catastrophic hurricanes the region faced its worst drought in record since at least 1950 (Herrera and Ault, 2017). Our group has called this event the “Pan-Caribbean drought” because it affected the entire Caribbean between 2013 and late 2016 in an unprecedented way. The Pan-Caribbean drought caused significant losses in agriculture and shortages in municipal water supply in most of the countries of the region (OCHA, 2015), and even put food security and energy generation at risk, especially in Haiti (OCHA, 2015). Furthermore, according to the UN Office for Coordination of Humanitarian Affairs (OCHA), 1 million people were affected by the drought in Haiti, and further exacerbated the situation of 3.8 million food-insecure people in the country. These statistics indicate that, although hurricane-driven disasters can cause significant and rapid damage, slow but persistent droughts can also wreak havoc in vulnerable countries like those of the Caribbean.

My own family struggled with the drought, as we rely on the rainy season to irrigate our crops and maintain our small farms in San Juan, located in the southwestern Dominican Republic (Fig. 1). Back September, my father actually asked me, “When will the rain come back here again?” He was hoping that—because I study drought—I might have insight into the seasonal outlook for his part of the world. Of course, this is not an easy question to answer, at least for now. During the drought, I talked with my family regularly and heard stories of the hardships faced by their neighbors and friends. A fellow farmer told my father that he and his family needed to move to San Domingo, at least temporarily, to find a job and pay the losses due to the drought. “It was very sad to leave the farm,” he said. Many other farmers were forced to take jobs transporting people and goods on motorcycles, a temporary occupation known as “motoconcho,” which translates roughly to “motor shell.”

Of course, when the drought finally ended, the eroded fields and bare soils were met with torrential rains from the hurricane season. The relatively rapid shift from drought to wet conditions in the Caribbean driven by Maria and Irma is evident by analyzing the regionally averaged Palmer Drought Severity Index (PDSI) time series (Fig. 2). The PDSI is a soil moisture proxy commonly used in the U.S. for drought monitoring and research, and it is calculated using precipitation and evapotranspiration as water supply and demand, respectively. As shown in Fig. 2, the Pan-Caribbean drought ended in December 2016 (PDSI > 0.5), but because of hurricanes Maria and Irma, the region shifted from near normal conditions (PDSI between > -0.5 and 3.0) in September 2017. This highlights the important role of tropical cyclones in suddenly ending droughts or worsening pluvials (a period of above-normal wetness, or just the opposite of a drought) in the Caribbean. Nevertheless, a very active hurricane season such as in 2005 or 2017, does not necessarily mean a very wet interval for the Caribbean, because this also depends on the track of tropical cyclones.

Figure 2: Regionally-averaged Palmer Drought Severity Index (PDSI) time series of Caribbean islands. The “Pan-Caribbean drought” ended in late 2016, although slight drought further built up in early 2017, and lasted until the very active hurricane season of the same year in the North Atlantic. We calculated the PDSI using precipitation data from the Global Precipitation Climatology Centre (GPCC), and temperature, wind speed, cloud cover, and vapor pressure from NCEP-NCAR reanalysis. Potential evapotranspiration (PET) was calculated using the FAO Reference PET as in Herrera and Ault, 2017.

Is this kind of rapid change unusual in the Caribbean?
Using our Caribbean drought atlas, we have found that a relatively rapid change from dry to wet––and even very wet––conditions is not uncommon in the Caribbean, at least when the situation involves tropical cyclones. For example, in August 2008 hurricane Gustav and tropical storm Hanna affected Hispaniola Island, and their heavy rainfall shifted the mild drought (PDSI = -1.8) in July to moderately wet conditions (PDSI = 2.9) in August. Similarly, major hurricanes David in August 1979 and Georges in September 1998 further diminished previous drier conditions in the Dominican Republic. Although the relatively short-life span of tropical cyclones, they usually produce significant amount of rain that eventually end drought and/or worsen pluvials in a relatively short period of time.

The highly complex topography in many Caribbean islands further plays a role in intensifying rainfall amounts due to tropical cyclones at very local scales. This means that, although a single hurricane can affect a whole island, its rainfall effects are not always homogeneous. For example, hurricane Flora hit the southwestern portion of Hispaniola Island in October 1963, but its heaviest rainfall was registered in a relatively small portion of the Island; specifically, 57 inches (~1448 mm) accumulated in Miragoane (southwestern Haiti, Roth, 2008). In fact, during hurricanes Irma and Maria, my family’s town did not experience the extensive flooding as in northern Dominican Republic, mostly because of the rain shadow formed by the main range of the country, the Cordillera Central, and the path of both hurricanes. “Here it is not raining that much at all… the rain is even beneficial for us to overcome the––remaining––drought here… one of my farm is still a kind of dry…”, said my father when I asked him about hurricane Irma effects in San Juan. In contrast, farmers in northeastern Dominican Republic were severely affected by record-breaking flooding due to Maria. This fact points out the need to have high-resolution gridded climate products in the Caribbean that resolves not only many of the tiny islands of the region, but also its complex topography.

Rapid hydroclimate changes might worsen in the Caribbean due to climate change
Hydroclimate extremes are expected to worsen in the future due to anthropogenic climate change in the Caribbean, especially drought (IPCC, 2014) (Fig. 3). According to the Intergovernmental Panel for Climate Change (IPCC) 5th assessment report, a significant increase in consecutive dry days might occur in the region, ranging from 10 to 25% more dry days (days with less than 1 mm of rain) (Fig. 3). In contrast, although maximum five-days precipitation is also expected to slightly increase in islands such as Cuba, climate models project a decrease in five-days precipitation in most of Caribbean islands (IPCC, 2014 Fig. 3). However, the projected change in extreme precipitation is insignificant, as compared to changes in the number of dry days (IPCC, 2014). Furthermore, even though climate models indicate a (non-significant) decrease in tropical cyclones frequency in North Atlantic, they also show a slight increase in precipitation rate (IPCC, 2014). In any case, an increment in the number of dry days, in addition to tropical cyclones with higher precipitation rates would make rapid hydroclimate changes more common in the Caribbean, and might pose a significant challenge for the region in terms of climate change adaptation capacity.

Future challenges
Understanding the causes of hydroclimate variability––especially relatively rapid changes––is critical to improve climate change adaptation capacity in the Caribbean. Given that currently the dynamical causes and characteristics of Caribbean hydroclimate are not well constrained (mostly due to the lack of high-quality and long-term climate data), an accurate forecast of rapid hydroclimate changes––as those related to drought-tropical cyclones––could be challenging. Nevertheless, comprehend this kind of changes is critical to improve the resilience to climate change of Caribbean nations. This is further crucial to ensure a sustainable future for the more than 50 million people projected to live in Caribbean the next decades, provided its insular nature and small geographic size.

Figure 3: Changes in global extreme precipitation (maximum five-day precipitation) and number of dry days from the IPCC 5th Assessment Report (2014). Although changes in extreme precipitation are not significant for the Caribbean, the number of dry days (

1. Daly, C., E. H. Helmer, M. Quiñonez (2003): Mapping the climate of Puerto Rico, Vieques and Culebra. Int. J. of Clim., 23, 1359-1381.
2. Herrera, D., and Ault, T. R. (2017): Insights from a new high-resolution drought atlas for the Caribbean spanning 1950-2016. J. of Clim., 30, 7801-7825.
3. Intergovernmental Panel for Climate Change (IPCC) (2014): Climate Change 2014: Synthesis Report. R.K. Pachauri and L. A. Meyer, Eds., IPCC, 151 pp.
4. National Oceanic and Atmospheric Administration (NOAA) (2017): NOAA Media Release (http://www.noaa.gov/media-release/extremely-active-2017-atlantic-hurrica...)
5. OCHA, 2015: Drought in Central America in 2015: Situation report (as of October 6, 2015). United Nations Office for the Coordination of Humanitarian Affairs (OCHA). [Available at: http://www.redhum.org/uploads/documentos/pdf/Sitrep_OCHAROLAC_Drought_in...
6. Roth, D. M., 2008: Hurricane Flora (1963) rainfall graphic. Tropical cyclone point maxima, National Weather Service. Accessed 9 June 2012. [Available online at http://www.wpc.ncep.noaa.gov/tropical/rain/flora1963filledrainblk.gif.]