A recent collaboration between A. Park Williams and Chris Funk has been published in the scientific journal Climate Dynamics. The work shows a connection between warming in the Indian Ocean and suppression of convective rainfall during the Long Rains season of March through June in eastern Kenya and Ethiopia. The paper clearly has food security implications, and it also has implications for the way we interpret projections of tropical circulation and precipitation made by GCMs. While GCMs tend to project a decreased atmospheric Walker Circulation over the tropical Pacific and Indian Oceans (slower over-turning circulation between the eastern tropical Pacific cold-tongue region and western Pacific/Indian Ocean Warm pool), the paper shows that, thus far, a slowed Walker circulation cannot be detected in the observed climate record. Instead, we show that the Warm Pool has extended westward into the Indian Ocean, causing the western, convective branch of the Walker Circulation to extend to the west as well. As evaporation, convection, and precipitation have increased over the Indian Ocean, circulation has been altered in surrounding areas including the Horn of Africa.
The article is published with Open Access and can be downloaded from SpringerLink here. Below, Park Williams answers some questions about the research.
Q: What is "new" about this research?
A: This research shows compelling evidence coming from a wide range of datasets that point toward a strong connection between warming in the central tropical Indian Ocean and drought during March-June in eastern Ethiopia and Kenya. This concept has been introduced by others, but this study uses a whole lot of data to build a more compelling case than has previously been offered. The details of the study are described below.
Q: What are the bottom line conclusions (as applicable to the general public)?
A: Bottom line -
We show a connection between warming in the Indian Ocean and suppression of convective rainfall during the Long Rains season of March through June in eastern Kenya and Ethiopia.
The story is about the Tropical Warm Pool: where the eastern tropical Pacific Ocean meets the western tropical Indian Ocean, in the region of the Indonesian and Malaysian Islands. The ocean's surface is warmer here than anywhere else on earth. The fact that the ocean's surface is very warm has two very important impacts within the Warm Pool region. One, water evaporates from the Warm Pool's surface very rapidly, causing the air above the Warm Pool to be very humid. Two, warm air rises, and where warm, humid air rises, you get heavy rain. When the humid air condenses into liquid water high above the Warm Pool, a massive amount of energy is released into the atmosphere. This energy causes wind to flow in all directions away from the area where the storm clouds are produced.
While the ocean's surface is very warm within the Warm Pool region, it is much cooler to the east and west, in the central Pacific Ocean and the central/western Indian Ocean, respectively. Much of the dry air traveling through the upper atmosphere away from the stormy Warm Cool region eventually sinks toward the Oceans' surface over these cool waters. Once near the surface, this air once again travels toward the Warm Pool region preparing for the cycle all over again by warming and picking up moisture along the way.
The Story -
As the globe has warmed over the last century, the Warm Pool region has warmed as well. As it has warmed, it has expanded. On its western side, the Warm Pool has expanded by about 4,000 km into the central Indian Ocean. This westward expansion of the Warm Pool has caused surface waters in central Indian Ocean to warm about one-and-a-half times faster than the global average temperature, and about three times faster than the cooler tropical waters of the central Pacific Ocean. Westward expansion of the Warm Pool has led to increased humidity, upward movement of air, and storm activity over the central and western Indian Ocean. As the region of warm water, convection, and storminess has expanded to the west, the western sinking branch of the circulation system has also moved to the west. During the March-June long-rains season in Kenya and Ethiopia, the rate at which dry air sinks over northern Africa has substantially increased, in step with storminess over the central Indian Ocean. When this dry air returns to the surface over northern Africa, much of it flows east, back to the Indian Ocean, and travels through the Horn of Africa on the way. This means trouble for Kenya and Ethiopia because the eastward flow of dry air from northern Africa works to block the moist westward winds that bring rain water onto land from the Indian Ocean. As a result of this change in atmospheric circulation over the Warm Pool and surrounding areas, long-rains precipitation has decreased by over 30% throughout much of eastern Ethiopia and Kenya. Because the expansion of the Warm Pool has been so well correlated with global temperatures, and because we believe that global temperatures are very likely to continue increasing for a long time to come, we anticipate that average March-June precipitation totals in Kenya and Ethiopia will continue decreasing or remain below the historical average for a long time.
Q: Why were the IPCC projections incorrect? Why did they project increased rainfall in eastern Africa? Why did your research show something different?
A: While state-of-the art global climate models generally do an impressive job of simulating spatial differences in average precipitation (dry areas are simulated to be dry, wet areas simulated to be wet), they don't do so well at simulating year-to-year variability in precipitation at many places on earth. In these areas, where models cannot accurately simulate year-to-year or decade-to-decade swings in precipitation, it is unwise to trust model forecasts of future precipitation. For example, models have a notoriously difficult time simulating year-to-year precipitation variability over tropical land masses. The shortcoming of modeled precipitation data is at least partly due to the fact that models still have a tough time accurately simulating El Niño, a dominant controller of precipitation throughout the tropics. Importantly, models tend to forecast a trend toward a more El Niño-like climate. This means increased precipitation over the central Pacific and decreased precipitation over the Warm Pool region. This also generally means increased precipitation in Ethiopia and Kenya during March through June. What we show, however, is mainly the opposite. Global warming has, at least so far, been strongly associated with substantially increased precipitation over the rapidly warming western edge of the Warm Pool region, in the central Indian Ocean. This strong correlation between Indian Ocean precipitation and global temperatures is not represented in any of the models used by the IPCC.
Q: How did you conduct the research? Can you explain this in as simple and basic terminology as possible? Also, make sure is clearly states why the USGS did this as opposed to NOAA since it seems to be research on atmospheric-oceanic interaction. I think you looked at historical datasets on the circulation and climate systems in the Indian and Pacific Oceans to see what drives variations and how that impacts rainfall in eastern Africa. Did you use satellite imagery, new research by the USGS and University of California, compile existing data, etc.?
A: We used many datasets of temperature, wind speed, and precipitation and found that among the many processes causing these variables to vary from year to year in the tropical Indian and Pacific Ocean region, the very most dominant mode of year-to-year variability during at least the last 60 years has been associated with the globe's average temperature. This means that the increasing global temperature has caused more climate variability in the tropical Indian and Pacific Ocean region than El Niño. Above the central Indian Ocean, sea-surface temperatures, wind speeds, humidity, vertical atmospheric rising motion, and precipitation are all strongly correlated with global temperatures. All of these factors work together to suppress March-June precipitation in Kenya and Ethiopia.
We did our analysis of Kenya and Ethiopia precipitation using a dataset compiled by Chris Funk of the USGS. Dr. Funk processed precipitation data collected at hundreds of weather stations in the Horn of Africa to calculate the regional March-June totals for Ethiopia and Kenya. The precipitation records agree fairly well with precipitation datasets calculated by other researchers, but Dr. Funk's precipitation dataset is thought to be more accurate because it incorporates data from many more stations.
The global climate data come from a variety of sources available to the public, mainly through the National Climate Data Center. We also accessed IPCC model data from the World Climate Research Programme Multi-Model Database.
The USGS is concerned with this work because it has large implications for food security in the Horn of Africa, a region prone to frequent droughts and very high levels of hunger. The division of the USGS specifically concerned with this research is the Famine Early Warning Systems Network (FEWS NET) project, funded by the U.S. Agency for International Development (USAID) Office of Food for Peace.
Q: Is increased drought projected for the future? Or was the future that not looked at in this study?
A: Yes, increased drought is projected for the future. While there are many factors that influence long-rains precipitation totals on a year to year basis, meaning there will continue to be occasional very wet years and occasional very dry years, we expect that increased sea-surface temperatures in the Indian Ocean will continue to decrease the average March-June precipitation totals in Ethiopia and Kenya. In addition, increasing temperatures will increase the evaporative demand of the atmosphere because water evaporates more quickly in warm air. Reduced precipitation in combination with warming is expected to put a major strain on agricultural systems.