Multiyear La Niña Events and Multiseason Drought in the Horn of Africa

Weston Anderson aEarth System Science Interdisciplinary Center, University of Maryland, College Park, College Park, Maryland
bNASA Goddard Space Flight Center, Greenbelt, Maryland

Search for other papers by Weston Anderson in
Current site
Google Scholar
PubMed
Close
https://orcid.org/0000-0003-3755-9943
,
Benjamin I. Cook cNASA Goddard Institute for Space Studies, New York, New York
dLamont-Doherty Earth Observatory, Palisades, New York

Search for other papers by Benjamin I. Cook in
Current site
Google Scholar
PubMed
Close
,
Kim Slinski aEarth System Science Interdisciplinary Center, University of Maryland, College Park, College Park, Maryland
bNASA Goddard Space Flight Center, Greenbelt, Maryland

Search for other papers by Kim Slinski in
Current site
Google Scholar
PubMed
Close
,
Kevin Schwarzwald eInternational Research Institute for Climate and Society, Palisades, New York

Search for other papers by Kevin Schwarzwald in
Current site
Google Scholar
PubMed
Close
,
Amy McNally bNASA Goddard Space Flight Center, Greenbelt, Maryland
fU.S. Agency for International Development, Washington, D.C.

Search for other papers by Amy McNally in
Current site
Google Scholar
PubMed
Close
, and
Chris Funk gClimate Hazards Center, University of California, Santa Barbara, Santa Barbara, California

Search for other papers by Chris Funk in
Current site
Google Scholar
PubMed
Close
Free access

Abstract

One of the primary sources of predictability for seasonal hydroclimate forecasts are sea surface temperatures (SSTs) in the tropical Pacific, including El Niño–Southern Oscillation. Multiyear La Niña events in particular may be both predictable at long lead times and favor drought in the bimodal rainfall regions of East Africa. However, SST patterns in the tropical Pacific and adjacent ocean basins often differ substantially between first- and second-year La Niñas, which can change how these events affect regional climate. Here, we demonstrate that multiyear La Niña events favor drought in the Horn of Africa in three consecutive seasons [October–December (OND), March–May (MAM), OND]. But they do not tend to increase the probability of a fourth season of drought owing to the sea surface temperatures and associated atmospheric teleconnections in the MAM long rains season following second-year La Niña events. First-year La Niñas tend to have both greater subsidence over the Horn of Africa, associated with warmer waters in the west Pacific that enhance the Walker circulation, and greater cross-continental moisture transport, associated with a warm tropical Atlantic, as compared to second-year La Niñas. Both the increased subsidence and enhanced cross-continental moisture transport favors drought in the Horn of Africa. Our results provide a physical understanding of the sources and limitations of predictability for using multiyear La Niña forecasts to predict drought in the Horn of Africa.

© 2023 American Meteorological Society. For information regarding reuse of this content and general copyright information, consult the AMS Copyright Policy (www.ametsoc.org/PUBSReuseLicenses).

Corresponding author: Weston Anderson, weston.anderson@nasa.gov

Abstract

One of the primary sources of predictability for seasonal hydroclimate forecasts are sea surface temperatures (SSTs) in the tropical Pacific, including El Niño–Southern Oscillation. Multiyear La Niña events in particular may be both predictable at long lead times and favor drought in the bimodal rainfall regions of East Africa. However, SST patterns in the tropical Pacific and adjacent ocean basins often differ substantially between first- and second-year La Niñas, which can change how these events affect regional climate. Here, we demonstrate that multiyear La Niña events favor drought in the Horn of Africa in three consecutive seasons [October–December (OND), March–May (MAM), OND]. But they do not tend to increase the probability of a fourth season of drought owing to the sea surface temperatures and associated atmospheric teleconnections in the MAM long rains season following second-year La Niña events. First-year La Niñas tend to have both greater subsidence over the Horn of Africa, associated with warmer waters in the west Pacific that enhance the Walker circulation, and greater cross-continental moisture transport, associated with a warm tropical Atlantic, as compared to second-year La Niñas. Both the increased subsidence and enhanced cross-continental moisture transport favors drought in the Horn of Africa. Our results provide a physical understanding of the sources and limitations of predictability for using multiyear La Niña forecasts to predict drought in the Horn of Africa.

© 2023 American Meteorological Society. For information regarding reuse of this content and general copyright information, consult the AMS Copyright Policy (www.ametsoc.org/PUBSReuseLicenses).

Corresponding author: Weston Anderson, weston.anderson@nasa.gov

1. Introduction

Drought has been a frequent trigger of acute food insecurity in the Horn of Africa (38°–53°E, 5°S–8°N) over the last decade (Anderson et al. 2021; Funk et al. 2018; Maxwell and Hailey 2020; Maxwell and Majid 2016; Shukla et al. 2021). As of February 2022, portions of the Horn of Africa are experiencing a multiseason drought in which rainfall has been below normal during the short rains (October–December) of 2020, the long rains (March–May) of 2021, and again in the short rains of 2021, which is contributing to the ongoing food security crisis in the region (FEWS NET 2021). In this context, there is considerable interest in the potential to understand and predict such multiyear drought events at lead times beyond the seasonal time scale, which would allow governments and humanitarian aid agencies to better prepare for the acute food insecurity caused by consecutive seasonal droughts.

One of the primary sources of predictability for seasonal hydroclimate forecasts are sea surface temperatures (SSTs) in the tropical Pacific, including El Niño–Southern Oscillation (ENSO) (Lenssen et al. 2020; Shukla et al. 2019; Funk et al. 2018). Operational seasonal forecasting systems, such as the North American Multi-Model Ensemble (NMME), demonstrate skill in predicting Niño-3.4 anomalies at leads up to 8–12 months in advance (Barnston et al. 2019, 2012). And while NMME forecasts are not available for leads longer than 12 months, machine learning-based multiyear forecasts of Niño-3.4 anomalies demonstrate skill up to 16 months in advance (Ham et al. 2019). Such forecasts of tropical Pacific SSTs are relevant for forecasting drought in East Africa during both the October–December and March–May season. Indeed, operational NMME precipitation forecasts are more skillful during strong ENSO years in East Africa (Shukla et al. 2019).

La Niña events in particular may be predictable multiple years in advance provided the correct initial conditions (Wu et al. 2021; DiNezio et al. 2017a,b; Luo et al. 2017). It has long been recognized that both El Niño and La Niña exhibit life cycles with predictable periods of growth and decay that are phase locked to the seasonal cycle (Zebiak and Cane 1987; Rasmusson and Carpenter 1982). Recently, it has been identified that La Niñas are more likely to persist for two years as compared to El Niños (Okumura and Deser 2010). This multiyear persistence of La Niña is particularly pronounced immediately following strong El Niños, when subsurface oceanic heat content is discharged off the equator, the thermocline in the tropical Pacific is strongly shoaled in the east Pacific, and the warming of the Atlantic and Indian Oceans acts to enhance the easterly winds in the central Pacific (DiNezio and Deser 2014; Wu et al. 2019). The cycle of strong El Niños being followed by multiyear La Niñas represents a signal that may be predictable up to two years in advance (Wu et al. 2021; DiNezio et al. 2017a,b; Luo et al. 2017), although to date operational forecasts are not regularly issued at these lead times.

Multiyear La Niñas are not only potentially predictable, but they are also detrimental to agriculture in the bimodal rainfall regions of East Africa, where La Niñas strongly favor drought during the short rainy season (October–December) and may also favor drought during the long rainy season (March–May). The negative phase of the Indian Ocean dipole and La Niña, which often occur together, favor drought during the short rains by strengthening the Indian Ocean branch of the Walker circulation (Liebmann et al. 2014; Liu et al. 2020; Blau and Ha 2020; Goddard and Graham 1999; Tierney et al. 2013; Dutra et al. 2013). Rainfall is suppressed under these conditions because convection and atmospheric ascent over the west Pacific is enhanced while over the western Indian Ocean and the Horn of Africa atmospheric descent intensifies and suppresses convection. During the March–May long rains, the connection between rainfall in the Horn of Africa and SSTs is more complex. Historically, ENSO has had no significant teleconnection to the long rains (Nicholson 2017), although recent research has demonstrated that a warm west Pacific and cold east/central Pacific can force drought in the Horn of Africa by strengthening the Walker circulation (Funk et al. 2019, 2018; Ummenhofer et al. 2018; Williams and Funk 2011; Hoell and Funk 2013, 2014; Liebmann et al. 2017; Lyon and DeWitt 2012; Funk and Hoell 2015) and modifying moisture fluxes into the region (Hoell and Funk 2013). The combination of La Niñas with warm west Pacific sea surface temperatures and the negative phase of the Indian Ocean dipole, therefore, has the potential to force consecutive years of drought (Hoell and Funk 2014) similar to the present multiyear drought in the region. In 2015/16, for example, a strong El Niño followed by consecutive La Niñas led to substantial food insecurity in eastern and southern Africa (Funk et al. 2018).

However, SST patterns in the tropical Pacific and adjacent ocean basins often differ substantially between first- and second-year La Niñas, which can change how these events affect regional climate. First-year La Niñas develop following El Niños while second-year La Niñas persist from the previous La Niña. Because of this, SSTs in the central Pacific tend to be colder during first-year events (Wu et al. 2019), while SSTs are warmer and the associated convective heating in the west Pacific is greater during the summer of a developing first-year La Niña as compared to that of a second-year La Niña (Jong et al. 2020). Tropical Atlantic SSTs differ as well, tending to be warmer during first-year La Niñas as compared to second-year La Niñas (Okumura et al. 2017) during November–April. These differences between first-year and second-year La Niñas have been shown to substantially affect atmospheric teleconnections during boreal winter (Okumura et al. 2017) when ENSO events tend to peak in strength, during boreal summer (Jong et al. 2020) when events are developing, and during boreal spring (Tokinaga et al. 2019) when events are decaying.

Here, we characterize how multiyear and single-year La Niñas affect the probability of consecutive seasons of drought in the Horn of Africa. We will address two main questions in our analysis: 1) How do single- and multiyear La Niñas affect drought occurrence during consecutive rainy seasons over East Africa? 2) How do the mechanisms differ across seasons during single- versus multiyear La Niña events? Improving our understanding of the dynamics of multiyear La Niñas as they relate to drought and water resources in food insecure regions will directly inform the potential usefulness of long-lead ENSO forecasts for these areas. We analyze precipitation due to its importance for rainfed cropping systems and runoff as an integrated metric of moisture supply to pastoral water points and river flow in the region.

2. Methods

For data on precipitation, we use both the Global Precipitation Climatology Center (GPCC) version 2020 (Schneider et al. 2018) as well as the Climate Research Unit (CRU) time series, version 4.05 (Harris et al. 2020), both gridded to 0.5° globally, from 1920 to 2019. We begin our analysis in 1920 because a greater number of precipitation gauge stations report data after this period, resulting in greater coverage over the Horn of Africa in the gridded GPCC product (Fig. S1 in the online supplemental material). For data on runoff, we use the observational-based Global Runoff Reconstruction dataset (GRUN; Ghiggi et al. 2019), which uses precipitation and temperature reanalyses to train a machine learning model to predict monthly runoff rates. To calculate precipitation and runoff anomalies we first aggregate values to October–December and March–May totals, then calculate differences from the long-term (1949–2019) average, before finally converting to a z score by dividing by the standard deviation over the same period. Note that using a z score of three-month seasonal precipitation anomalies is mathematically equivalent to the commonly used three-month standardized precipitation index (SPI). To diagnose circulation anomalies in the post-1950 period, we use the ERA5 global reanalysis (Hersbach et al. 2020) because the precipitation from this reanalysis has been shown to be more accurate in the Horn of Africa as compared to previous reanalyses (Gleixner et al. 2020). We use atmospheric winds and the vertical integral of water vapor flux, which is available as a standard output for ERA5 calculated as the moisture flux in a column of air extending from the surface of Earth to the top of the atmosphere. We calculate anomalies as differences from the 1950–2020 long-term average.

For data on SST, we use values from the Hadley Centre Sea Ice and Sea Surface Temperature dataset (HadISST) (Rayner et al. 2003). Because there have been significant trends in SST over the 1920–2019 period, we detrend SST values locally by subtracting the low-frequency filtered tropical and subtropical mean SST value. We first calculate the trend by calculating the area-weighted SST value in the 33°S–33°N region before smoothing the resulting time series with a 30-yr low-pass locally weighted scatterplot smoothing (LOWESS) filter and subtracting this trend from each local SST value. We then remove the average SST value at each point to calculate anomalies before aggregating to seasonal October–December and March–May values.

We identify La Niña events as years when SST anomalies in the Niño-3.4 region are at least −0.5°C during October–December. We use the October–December season to identify events because ENSO events tend to peak during boreal winter and because October–December is the short rains season, which ensures the identified events had developed by the time of the short rains. We next separate La Niña events into multiyear La Niñas and single-year La Niñas using the same −0.5°C threshold. In this analysis we do not consider the small number of third-year La Niña events on record. While the contemporaneous relationship between March and May Niño-3.4 and East African rains is weak (Funk et al. 2018), there have been frequent dry East African long rains following October–December La Niña events.

From this point forward we will refer to the first year of a La Niña, regardless of whether it eventually developed into a second-year event, as a “first-year La Niña” with the associated seasons being “October–December of year one” or “March–May of year one” to refer to the short rains coincident with the peak of the event or the long rains during the decaying phase of an event. La Niña events that do not develop into a second-year La Niña will be referred to as “single La Niñas,” while those that do develop into a second consecutive event will be referred to as “double La Niñas.” The October–December season coincident with the peak of the second year of a double La Niña will be referred to as “October–December of year two” with the following March–May being referred to as “March–May of year two.”

To assess the statistical significance of La Niña composites of SST, precipitation, and runoff, we use a double bootstrap superposed epoch analysis, as in Rao et al. (2019). A double bootstrap superposed epoch analysis allows us to evaluate the statistical significance of the La Niña signal while accounting for sample size of our La Niña event set. This approach simultaneously evaluates two questions: 1) How robust is an apparent drought signal within the event set of La Niña years? 2) How likely are we to observe an equally strong drought signal by drawing the same number of years at random from the full record of all years? The first question addresses the possibility that the observed signal is the result of outliers within the event set of La Niña years, while the second question addresses the possibility that the observed signal would arise by chance alone given the observed number of La Niña years. We first block resample each variable in contiguous three year blocks 1000 times with replacement from the full 1920–2019 period to create the full “noise” distribution against which the composite mean will be evaluated. We then resample from each La Niña composite 1000 times, drawing 75% of events without replacement before calculating the mean of the resampled composite. The subsequent distribution of composite means represents the La Niña event distribution. For each noise and each event distribution we plot the 5th, 10th, 20th, and 80th, 90th, and 95th percentiles.

3. Results

Twenty-five La Niñas were identified from 1920 to 2019 (Figs. 1a,b). Seventeen of these were first-year La Niñas, of which eight were followed by at least one consecutive year of La Niña. The nine first-year La Niña events that did not develop into a second year of La Niña were 1924, 1933, 1938, 1942, 1945, 1964, 1988, 1995, and 2007. The eight first-year La Niñas that were followed by a second consecutive La Niña were 1949/50, 1954/55, 1970/71, 1973/74, 1983/84, 1998/99, 2010/11, and 2016/17. Here the year refers to the October–December near the peak of each event.

Fig. 1.
Fig. 1.

(a),(b) Sea surface temperature anomalies in the Niño-3.4 region, (c),(d) precipitation anomalies in the Horn of Africa, and (e),(f) runoff anomalies in the Horn of Africa during (left) first-year and second-year events of multiyear La Niñas and (right) single-year La Niñas. Individual events shown with dotted lines and composite means shown with solid lines. The Horn of Africa is defined as the region between 38° and 53°E and between 5°S and 8°N.

Citation: Journal of Hydrometeorology 24, 1; 10.1175/JHM-D-22-0043.1

Comparing the October–December SSTs of the first year of double La Niñas to those of single-year La Niñas (Figs. 1a,b) indicates that the former tended to develop from stronger El Niños but tended to be similar in their peak intensity. Double La Niñas were persistently colder during the spring and summer between peak event intensities, often not warming above −0.5°C during the months between La Niña peaks.

We found that both single and double La Niñas affect precipitation and subsequently runoff during the October–December short rains in the Horn of Africa, consistent with past literature indicating the strong influence of ENSO during the short rains season in the Horn of Africa (Liebmann et al. 2014; Liu et al. 2020; Blau and Ha 2020; Goddard and Graham 1999; Tierney et al. 2013; Dutra et al. 2013). All three kinds of La Niña years–single La Niñas, the first year of double La Niñas, and the second year of double La Niñas–tend to be associated with below normal precipitation and runoff in October–December (Figs. 1c–f; see also supplemental Fig. S2). The strength of the reduction in precipitation and runoff is greater for the first-year La Niñas as might be expected from the greater intensity of these events as compared to second-year La Niñas, which would more strongly modify the Walker circulation, resulting in more strongly suppressed convection in the region. The strength of the precipitation response is slightly greater during first-year La Niñas as compared to that of single-year La Niñas, although the reasons for this are unclear.

The March–May long rains during the decaying phase of La Niñas demonstrated a greater diversity of precipitation and runoff responses. The March–May long rains during the decaying phases of both types of first-year La Niñas were below normal, while the March–May long rains following second-year La Niñas tended to be wetter than normal (Figs. 1c,d; supplemental Fig. S2).

Whether these results are robust, however, depends on both the variability of events within each La Niña composite and the variance of the years not included in the composite. We therefore assess the robustness of event composites by using a double-bootstrap superposed epoch analysis (Rao et al. 2019), which subsamples both the La Niña composite itself and the years not included in the La Niña composite to generate confidence estimates around the expected background “noise” of events not included in the composite as well as a confidence estimate on the La Niña composite mean.

During the short rains season [October–December (OND)] season, all three types of La Niñas have a distribution of SST composite means that are colder than the 95% confidence interval of the background noise of SSTs (Figs. 2a,b). This is to be expected given that the OND season is the season we use to define the events that are included in the composites. While the first year of a double La Niña tends to be colder than both a second-year and a single-year La Niña, the average composite mean of a first-year event is not outside of the 95% confidence interval of either a second-year or single-year La Niña.

Fig. 2.
Fig. 2.

Uncertainty around composite means from Fig. 1 estimated using a double-bootstrap superposed epoch analysis. Shading shows the 5th, 10th, 20th, and 80th, 90th, and 95th percentiles of the composite mean, while the solid lines are composite means, as in Fig. 1 for (a),(b) sea surface temperature anomalies in the Niño-3.4 region, (c),(d) precipitation anomalies in the Horn of Africa, and (e),(f) runoff anomalies in the Horn of Africa during (left) first-year and second-year events of multiyear La Niñas and (right) single-year La Niñas. Dotted lines represent bootstrapped uncertainty estimates as the 5th and 95th (black), 10th and 90th (gray), or 20th and 80th (light gray) percentiles of composite means calculated based on years drawn at random from the full record (see methods).

Citation: Journal of Hydrometeorology 24, 1; 10.1175/JHM-D-22-0043.1

SSTs during the March–May (MAM) season, on the other hand, are more different from one another. SSTs remain significantly lower than the background variability (e.g., the 95% confidence interval of the composite mean does not overlap with that of the background variability) after the first year of double La Niñas. This indicates that the tropical Pacific remains significantly colder than normal from March through the end of May between La Niña peaks. SSTs after second-year and single-year La Niña events tend to return to neutral more quickly such that by May the 95% confidence interval of SSTs are within the background 95% confidence interval of the noise distribution in both cases.

October–December precipitation and runoff are lower than normal during first-year La Niñas of both types, although double La Niña are slightly drier than single-year La Niñas. During the first year of a double La Niña, October–December precipitation and runoff is near the 5th percentile, while during the first year of a single La Niña precipitation and runoff is near the 20th percentile. October–December rainfall during second-year La Niñas (Fig. 2c) also tends to be drier than normal on average, although there is a greater event-to-event difference and they tend to be less dry than first-year events. During second-year La Niñas, the 95% confidence interval of October–December precipitation and runoff is significantly wider than that of first-year La Niñas. It is generally below zero but not below the 5th percentile of the noise distribution, and the composite mean is not drier than the 20th percentile. These results are similar using CRU time series (TS), version 4.05, as an alternative gridded precipitation product (supplemental Fig. S3).

The March–May long rains season tends to be dry following first-year events of both types but wet following second-year La Niñas. While the 95% confidence interval of March–May precipitation during first-year La Niñas overlaps with the 20th percentile of the noise distribution, that of second-year March–May precipitation is centered at or above zero (Fig. 2c), depending on the month. This illustrates that not only does March–May season following second-year La Niñas tend to be wetter than that following single-year La Niñas, it also tends to be wetter than average.

While Figs. 1 and 2 describe the time-evolution of regional-average precipitation over the Horn of Africa, Figs. 3 and 4 illustrate the spatial precipitation response. During a first-year La Niña, precipitation deficits are widespread during the October–December short rains and subsequent March–May long rains, although deficits are generally stronger during the first year of a double La Niña. Precipitation deficits during the October–December short rains of a second-year La Niña are more spatially confined to southeastern Kenya and southern Somalia but are not statistically significant. During the March–May long rains following a second-year event, virtually the entire Horn of Africa experiences normal or above-normal rainfall (Figs. 4g,h), although anomalies are again not statistically significant anywhere in the study domain. Individual years for first-year and second-year La Niñas are shown in supplemental Figs. S4 and S5. Overall, the individual years are heterogeneous, with several extremely wet years (1951, 2018), one exceptionally dry year (2000), and many with mixed anomaly patterns.

Fig. 3.
Fig. 3.

Average (a),(c) precipitation and (b),(d) runoff z scores in the Horn of Africa during single-year La Niñas during the OND short rains and the subsequent MAM long rains. Gray regions indicate climatologically dry areas, defined as receiving less than 30% of annual precipitation during the plotted months.

Citation: Journal of Hydrometeorology 24, 1; 10.1175/JHM-D-22-0043.1

Fig. 4.
Fig. 4.

Average (left) precipitation and (right) runoff z scores in the Horn of Africa during multiyear La Niñas (a)–(d) during the OND short rains of first-year events and the subsequent the MAM long rains during first-year events, as well as (e)–(h) during the OND short rains of second-year events and the subsequent the MAM long rains during second-year events. Gray regions indicate climatologically dry areas, defined as receiving less than 30% of annual precipitation during the plotted months.

Citation: Journal of Hydrometeorology 24, 1; 10.1175/JHM-D-22-0043.1

The different drought responses during the October–December short rains during each La Niña type is consistent with observed differences in the Indian Ocean SSTs. During October–December, the influence of the tropical Pacific is mediated by the Indian Ocean (Goddard and Graham 1999). In particular, the negative phase of the Indian Ocean dipole (IOD)–e.g., warm SSTs in the eastern tropical Indian Ocean and cool SSTs in the western tropical Indian Ocean–acts to reduce precipitation across the Horn of Africa (Saji and Yamagata 2003) due to a suppression of the Indian Ocean branch of the Walker circulation (Behera et al. 2005). During October–December of single-year La Niñas and the first year of double La Niñas, cold tropical Pacific SSTs co-occur with warm western Pacific SSTs and the negative phase of the IOD (Figs. 5a,c), resulting in widespread and statistically significant reductions in precipitation (Figs. 3a and 4a) and runoff (Figs. 3b and 4b). Both the warm western Pacific SSTs and the negative phase of the IOD act to weaken the Indian Ocean branch of the Walker circulation. During October–December of second-year La Niñas, however, neither the warm western Pacific SSTs nor the negative IOD are present, resulting in dry but statistically insignificant reductions in precipitation and runoff (Figs. 4e,f) despite the presence of La Niña conditions in the tropical Pacific (Figs. 2a and 5e).

Fig. 5.
Fig. 5.

Sea surface temperature anomalies during the (left) short rains and (right) long rains of (a),(b) single-year La Niñas as well as during the (c),(d) first year and (e),(f) second year of multiyear La Niñas. Hatching indicates SST anomalies that are not significant at the p < 0.9 level.

Citation: Journal of Hydrometeorology 24, 1; 10.1175/JHM-D-22-0043.1

The reasons for the discrepancy between March and May rainfall responses following first-year events of both types as compared to second-year La Niñas is not immediately obvious. It is true that the tropical Pacific SST in the Niño-3.4 region during the March–May of a first-year event in a double La Niña tends to be cooler than that of a single-year event or than that of a second-year event. But that does not explain why the March–May rainfall and runoff following a single-year La Niña is drier than that following the second year of a double La Niña, both of which have similar SST values in the Niño-3.4 region (Figs. 2a,b). The double-bootstrap superposed epoch analysis indicates that these differences in rainfall and runoff are unlikely to have occurred by chance. Given this, and provided that there is some indication that all three ocean basins may affect the rainfall in the Horn of Africa long rains, we first investigate the differences in SST anomalies outside of the Niño-3.4 region before linking these SST anomalies to the observed pattern of drought.

Tropical SST anomalies differ considerably during March–May following single-year La Niñas, during the first-year of double La Niñas, and during the second year of double La Niñas. During March–May after the first-year of double La Niñas (Fig. 5d), the SST in the tropical Pacific is considerably colder than during either single-year events (Fig. 5b) or second-year events (Fig. 5f). Both single-year events and the first-year of double La Niñas, however, tend to have cold western Indian Ocean SSTs, warm tropical west Pacific SSTs, and warm tropical Atlantic SSTs. The SST anomalies in the tropical Atlantic and west Pacific during the March–May following a second-year event, however, are relatively cooler and not statistically significant. Based on past literature, the differences in Atlantic SST anomalies have the potential to favor drought by enhancing cross-continental moisture transport (Camberlin and Okoola 2003; Okoola 1999a,b), while strong zonal SST gradients in the Pacific and warm west Pacific SSTs are likewise conducive to drought due to an enhanced Walker circulation (Funk et al. 2019, 2018; Ummenhofer et al. 2018; Williams and Funk 2011; Hoell and Funk 2013, 2014; Liebmann et al. 2017; Lyon and DeWitt 2012). And while the zonal overturning cell associated with the Walker circulation is less coherent during MAM than OND (Hastenrath et al. 2011), the clear impact of double La Niñas on March–May precipitation strength suggests a possible role for the basinwide vertical circulation. To assess whether these mechanisms support the enhanced probability of drought during first-year as compared to second-year La Niñas, we create composites for winds (Fig. 6) and vertically integrated moisture transport (Fig. 7) using ERA5 data from 1950 to the present.

Fig. 6.
Fig. 6.

Atmospheric circulation (a) climatology and (b) anomalies averaged from 5°S to 5°N during single-year La Niñas, as well as during the (c) first year and (d) second year of multiyear La Niñas.

Citation: Journal of Hydrometeorology 24, 1; 10.1175/JHM-D-22-0043.1

Fig. 7.
Fig. 7.

Vertically integrated moisture transport—defined as moisture transport integrated from the surface of Earth to the top of the atmosphere—(a),(b) climatology and (c),(d) anomalies during single-year La Niñas as well as during the (e),(f) first year and (g),(h) second year of multiyear La Niñas. Shading in each plot shows the magnitude of transport. In (c)–(h), positive values indicate transport that enhances the mean-state transport while negative values indicate moisture transport that is in the opposite sense of the mean-state transport. (left) All Africa domain; (right) the same quantities over East Africa.

Citation: Journal of Hydrometeorology 24, 1; 10.1175/JHM-D-22-0043.1

During first-year events of both types, atmospheric ascent is stronger than normal over the warmer than normal Atlantic and west Pacific SSTs, while over the Horn of Africa there is stronger than normal descent in the middle and upper atmosphere (Figs. 6a,c). Second-year La Niña events, on the other hand, tended to have enhanced subsidence over the tropical Atlantic rather than enhanced ascent, as well as midtropospheric ascent over the Horn of Africa. And while the March–May following second-year La Niña events do tend to show slightly enhanced ascent over the west Pacific, it is weaker than during first-year La Niñas and the subsequent anomalous descent tends to be over the eastern tropical Indian Ocean near 80°–100°E rather than over the Horn of Africa. An eastward-shifted subsidence would be less effective at suppressing convection over the Horn of Africa, consistent with the observed wetter conditions during second-year La Niña events. This result is consistent with that of Hoell et al. (2014), who find that La Niñas that occur with warmer west Pacific SSTs tend to shift the descending branch of the Walker circulation farther west over the Horn of Africa, while those with cooler west Pacific conditions are associated with descent over the Indian Ocean instead. That the long rains following second-year La Niñas tend to be wetter than those following first-year La Niñas, therefore, may be partially the result of a Walker circulation that is shifted farther westward during first-year events as compared to second-year events.

Figure 7 illustrates the climatological and anomalous vertically integrated moisture transport during the March–May long rains following each type of La Niña. Moisture is transported from the southeast into the region before being transported to the west across the continent either over northern Tanzania or through the Turkana Channel (Figs. 7a,b). During all La Niñas, more moisture than normal is transported into the region from the southeast and more moisture is also diverged out of the region to the northeast. The result of this is that there is little difference between first-year La Niñas and second-year La Niñas in terms of the vertically integrated moisture transport through the north, south, or east boundaries of the region (Fig. 7, calculation not shown). The major difference between the types of La Niñas is in the amount of moisture moved westward across the continent. During first-year events either a normal or greater-than-normal amount of moisture is moved westward through the Turkana Channel and over northern Tanzania out of the region, while during second-year events less moisture than normal is diverged out of the region across the continent, resulting in a greater net divergence of moisture out of the region during first-year events as compared to second-year events (Figs. 7c–h). To test how robust the relationship between tropical South Atlantic SSTs and cross-continental moisture transport is, we created similar composites based on tropical South Atlantic SST anomalies during ENSO neutral March–May years, which showed that a warm tropical South Atlantic tends to be associated with greater east-to-west moisture transport as compared to years with a cool tropical South Atlantic (supplemental Fig. S6). These results agree with past analyses that observe greater cross-continental moisture transport and surface divergence over East Africa when the tropical South Atlantic is warm and the Indian Ocean is cold (Camberlin and Okoola 2003; Okoola 1999a,b).

4. Discussion

We demonstrate that multiyear La Niña events favor drought in the Horn of Africa for three consecutive seasons (OND–MAM–OND), but that they do not tend to increase the probability of a fourth season of drought owing to the different sea surface temperatures and associated atmospheric teleconnections in the MAM long rains season following second-year La Niña events as compared to those following first-year La Niña events. During the long rains season following first-year La Niña events, the ascending branches of the Walker circulation are intensified over relatively warmer waters in the tropical Atlantic and the west Pacific, while subsidence is increased over the Horn of Africa, which contributes to the observed dry conditions in those years. Second-year La Niña events, on the other hand, are associated with weaker ascent near 140°E and enhanced atmospheric descent over the Indian Ocean rather than over the Horn of Africa and as a result the long rains are relatively wetter. The warmer tropical South Atlantic during first-year La Niñas as compared with second-year La Niñas is associated with increased cross-continental moisture transport out of East Africa and drier conditions in the Horn of Africa.

Atmospheric teleconnections during the long rains following first-year La Niñas are distinct from those following second-year La Niñas, therefore, due to differing sea surface temperatures throughout the tropics that affect the associated teleconnections. We note, however, that there exists substantial heterogeneity among second-year La Niña events in terms of both SST patterns and subsequent precipitation anomalies throughout the Horn of Africa. For example, second-year La Niñas have been associated with drought recently in 2000 and 2012 (Fig. S4). It is for this reason that we stress the importance of the SST patterns that are associated with drought conditions rather than the distinction between first-year and second-year La Niñas alone.

Our analysis is observationally based, which has the benefit of circumventing persistent model deficiencies in simulating both the mean state and variability of rainfall in East Africa (Yang et al. 2014, 2015). Our observational approach, however, is limited by a relatively small sample size. We identified only 25 La Niña events over the entire observational record, of which 16 were part of a double La Niña and nine were single La Niñas. Our conclusions, therefore, would benefit from being further evaluated in a modeling environment provided sufficient model skill in both the long rains and short rains seasons. In particular we highlight that a modeling environment could clarify the relative importance of a westward-shifted Walker circulation as compared to an enhanced cross-continental moisture transport for drying East Africa during the long rains following first- and second-year La Niña events.

When considering these results, and future potential outcomes, it is important to note that the dramatic long-term drying of March–May rains in the Horn of Africa has been linked to ongoing warming in the west Pacific compared to the central and east Pacific (Funk et al. 2019, 2018; Williams and Funk 2011; Lyon and DeWitt 2012; Funk and Hoell 2015). The observed strengthening of zonal SST gradients in the Pacific is consistent with a response to anthropogenic forcing known as the ocean dynamical thermostat mechanism (Clement et al. 1996; Seager et al. 2019). SST patterns favoring drought in the Horn of Africa may, therefore, continue to occur more frequently in the future compared to the first half of the twentieth century. Likewise, we may reasonably expect multiseason droughts to continue occurring at the relatively higher rates seen in the last 20 years (Hoell and Funk 2014), ensuring that they remain a major source of climate risk in East Africa. Past research has demonstrated that by using these zonal SST gradients in the Pacific droughts in March–May can be predicted as early as January (Funk et al. 2014). Such forecasts were developed using a statistical approach based on SST indices (Funk et al. 2014) and using climate analogs (Shukla et al. 2014), both of which demonstrate greater skill in March–May than the coupled dynamical climate model-based forecasts of the time. Improving our understanding of and our ability to predict these droughts should be a priority for adapting to anthropogenic climate change.

Our results provide physical understanding of the sources and limitations of predictability stemming from multiyear La Niña forecasts in the Horn of Africa. Such forecasts would be of considerable value for predicting acute food insecurity given the climate sensitivity of livelihoods in the region. The majority of the areas that we identify as experiencing multiyear drought during multiyear La Niña events are pastoral or agropastoral livelihood zones. Agropastoralists in this region already use seasonal forecasts to inform planting and harvest decisions (Luseno et al. 2003), while nomadic pastoralist households often rely upon food aid during times of drought (Rufino et al. 2013). Multiseason forecasts would allow for the coordination of food aid policies that often take months to complete, such as the direct shipment of grain to a region, as well as allowing agropastoral households to consider a wider variety of management strategies.

Acknowledgments.

We thank Andy Hoell and Laura Harrison for helpful conversations early in the development of this analysis. This publication was made possible through the support of the Bureau of Humanitarian Assistance, U.S. Agency for International Development, under the terms of PAPA AID-FFP-T-17-00001 Famine Early Warning Systems Network (FEWS NET).

Data availability statement.

All data used in this article are analyses of existing, publicly available datasets. GPCC and CRU precipitation data can be found at https://psl.noaa.gov/data/gridded/data.gpcc.html and https://crudata.uea.ac.uk/cru/data/hrg/, respectively. ERA5 data are available from https://cds.climate.copernicus.eu/, HadISST data are from https://www.metoffice.gov.uk/hadobs/hadisst/, and GRUN runoff data are from https://doi.org/10.6084/m9.figshare.9228176.

REFERENCES

  • Anderson, W., and Coauthors, 2021: Violent conflict exacerbated drought-related food insecurity between 2009 and 2019 in sub-Saharan Africa. Nat. Food, 2, 603615, https://doi.org/10.1038/s43016-021-00327-4.

    • Search Google Scholar
    • Export Citation
  • Barnston, A. G., M. K. Tippett, M. L. L’Heureux, S. Li, and D. G. DeWitt, 2012: Skill of real-time seasonal ENSO model predictions during 2002–11: Is our capability increasing? Bull. Amer. Meteor. Soc., 93, 631651, https://doi.org/10.1175/BAMS-D-11-00111.1.

    • Search Google Scholar
    • Export Citation
  • Barnston, A. G., M. K. Tippett, M. Ranganathan, and M. L. L’Heureux, 2019: Deterministic skill of ENSO predictions from the North American multimodel ensemble. Climate Dyn., 53, 72157234, https://doi.org/10.1007/s00382-017-3603-3.

    • Search Google Scholar
    • Export Citation
  • Behera, S. K., J.-J. Luo, S. Masson, P. Delecluse, S. Gualdi, A. Navarra, and T. Yamagata, 2005: Paramount impact of the Indian Ocean dipole on the East African short rains: A CGCM study. J. Climate, 18, 45144530, https://doi.org/10.1175/JCLI3541.1.

    • Search Google Scholar
    • Export Citation
  • Blau, M. T., and K.-J. Ha, 2020: The Indian Ocean dipole and its impact on East African short rains in two CMIP5 historical scenarios with and without anthropogenic influence. J. Geophys. Res. Atmos., 125, e2020JD033121, https://doi.org/10.1029/2020JD033121.

    • Search Google Scholar
    • Export Citation
  • Camberlin, P., and R. Okoola, 2003: The onset and cessation of the “long rains” in eastern Africa and their interannual variability. Theor. Appl. Climatol., 75, 4354, https://doi.org/10.1007/s00704-002-0721-5.

    • Search Google Scholar
    • Export Citation
  • Clement, A. C., R. Seager, M. A. Cane, and S. E. Zebiak, 1996: An ocean dynamical thermostat. J. Climate, 9, 21902196, https://doi.org/10.1175/1520-0442(1996)009<2190:AODT>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • DiNezio, P. N., and C. Deser, 2014: Nonlinear controls on the persistence of La Niña. J. Climate, 27, 73357355, https://doi.org/10.1175/JCLI-D-14-00033.1.

    • Search Google Scholar
    • Export Citation
  • DiNezio, P. N., C. Deser, Y. Okumura, and A. Karspeck, 2017a: Predictability of 2-year La Niña events in a coupled general circulation model. Climate Dyn., 49, 42374261, https://doi.org/10.1007/s00382-017-3575-3.

    • Search Google Scholar
    • Export Citation
  • DiNezio, P. N., and Coauthors, 2017b: A 2 year forecast for a 60–80% chance of La Niña in 2017–2018. Geophys. Res. Lett., 44, 11 62411 635, https://doi.org/10.1002/2017GL074904.

    • Search Google Scholar
    • Export Citation
  • Dutra, E., L. Magnusson, F. Wetterhall, H. L. Cloke, G. Balsamo, S. Boussetta, and F. Pappenberger, 2013: The 2010–2011 drought in the Horn of Africa in ECMWF reanalysis and seasonal forecast products. Int. J. Climatol., 33, 17201729, https://doi.org/10.1002/joc.3545.

    • Search Google Scholar
    • Export Citation
  • FEWS NET, 2021: The eastern Horn of Africa faces an exceptional prolonged and persistent agro-pastoral drought sequence. ICPAC/FEWS NET/FAO GIEWS/WFP/JRC Tech. Rep., 8 pp., https://mars.jrc.ec.europa.eu/asap/files/special_focus_2021_11.pdf.

  • Funk, C., and A. Hoell, 2015: The leading mode of observed and CMIP5 ENSO-residual sea surface temperatures and associated changes in Indo-Pacific climate. J. Climate, 28, 43094329, https://doi.org/10.1175/JCLI-D-14-00334.1.

    • Search Google Scholar
    • Export Citation
  • Funk, C., A. Hoell, S. Shukla, I. Blade, B. Liebmann, J. B. Roberts, F. R. Robertson, and G. Husak, 2014: Predicting East African spring droughts using Pacific and Indian Ocean sea surface temperature indices. Hydrol. Earth Syst. Sci., 18, 49654978, https://doi.org/10.5194/hess-18-4965-2014.

    • Search Google Scholar
    • Export Citation
  • Funk, C., and Coauthors, 2018: Examining the role of unusually warm Indo-Pacific sea-surface temperatures in recent African droughts. Quart. J. Roy. Meteor. Soc., 144, 360383, https://doi.org/10.1002/qj.3266.

    • Search Google Scholar
    • Export Citation
  • Funk, C., and Coauthors, 2019: Examining the potential contributions of extreme “Western V” sea surface temperatures to the 2017 March–June East African drought. Bull. Amer. Meteor. Soc., 100 (1), S55S60, https://doi.org/10.1175/BAMS-D-18-0108.1.

    • Search Google Scholar
    • Export Citation
  • Ghiggi, G., V. Humphrey, S. I. Seneviratne, and L. Gudmundsson, 2019: GRUN: An observation-based global gridded runoff dataset from 1902 to 2014. Earth Syst. Sci. Data, 11, 16551674, https://doi.org/10.5194/essd-11-1655-2019.

    • Search Google Scholar
    • Export Citation
  • Gleixner, S., T. Demissie, and G. T. Diro, 2020: Did ERA5 improve temperature and precipitation reanalysis over East Africa? Atmosphere, 11, 996, https://doi.org/10.3390/atmos11090996.

    • Search Google Scholar
    • Export Citation
  • Goddard, L., and N. E. Graham, 1999: Importance of the Indian Ocean for simulating rainfall anomalies over eastern and southern Africa. J. Geophys. Res., 104, 19 09919 116, https://doi.org/10.1029/1999JD900326.

    • Search Google Scholar
    • Export Citation
  • Ham, Y.-G., J.-H. Kim, and J.-J. Luo, 2019: Deep learning for multi-year ENSO forecasts. Nature, 573, 568572, https://doi.org/10.1038/s41586-019-1559-7.

    • Search Google Scholar
    • Export Citation
  • Harris, I., T. J. Osborn, P. Jones, and D. Lister, 2020: Version 4 of the CRU TS monthly high-resolution gridded multivariate climate dataset. Sci. Data, 7, 109, https://doi.org/10.1038/s41597-020-0453-3.

    • Search Google Scholar
    • Export Citation
  • Hastenrath, S., D. Polzin, and C. Mutai, 2011: Circulation mechanisms of Kenya rainfall anomalies. J. Climate, 24, 404412, https://doi.org/10.1175/2010JCLI3599.1.

    • Search Google Scholar
    • Export Citation
  • Hersbach, H., and Coauthors, 2020: The ERA5 global reanalysis. Quart. J. Roy. Meteor. Soc., 146, 19992049, https://doi.org/10.1002/qj.3803.

    • Search Google Scholar
    • Export Citation
  • Hoell, A., and C. Funk, 2013: The ENSO-related west Pacific sea surface temperature gradient. J. Climate, 26, 95459562, https://doi.org/10.1175/JCLI-D-12-00344.1.

    • Search Google Scholar
    • Export Citation
  • Hoell, A., and C. Funk, 2014: Indo-Pacific sea surface temperature influences on failed consecutive rainy seasons over eastern Africa. Climate Dyn., 43, 16451660, https://doi.org/10.1007/s00382-013-1991-6.

    • Search Google Scholar
    • Export Citation
  • Hoell, A., C. Funk, and M. Barlow, 2014: The regional forcing of Northern Hemisphere drought during recent warm tropical west Pacific Ocean La Niña events. Climate Dyn., 42, 32893311, https://doi.org/10.1007/s00382-013-1799-4.

    • Search Google Scholar
    • Export Citation
  • Jong, B.-T., M. Ting, R. Seager, and W. B. Anderson, 2020: ENSO teleconnections and impacts on U.S summertime temperature during a multiyear La Niña life cycle. J. Climate, 33, 60096024, https://doi.org/10.1175/JCLI-D-19-0701.1.

    • Search Google Scholar
    • Export Citation
  • Lenssen, N. J., L. Goddard, and S. Mason, 2020: Seasonal forecast skill of ENSO teleconnection maps. Wea. Forecasting, 35, 23872406, https://doi.org/10.1175/WAF-D-19-0235.1.

    • Search Google Scholar
    • Export Citation
  • Liebmann, B., and Coauthors, 2014: Understanding recent eastern Horn of Africa rainfall variability and change. J. Climate, 27, 86308645, https://doi.org/10.1175/JCLI-D-13-00714.1.

    • Search Google Scholar
    • Export Citation
  • Liebmann, B., and Coauthors, 2017: Climatology and interannual variability of boreal spring wet season precipitation in the eastern Horn of Africa and implications for its recent decline. J. Climate, 30, 38673886, https://doi.org/10.1175/JCLI-D-16-0452.1.

    • Search Google Scholar
    • Export Citation
  • Liu, W., K. H. Cook, and E. K. Vizy, 2020: Influence of Indian Ocean SST regionality on the East African short rains. Climate Dyn., 54, 49915011, https://doi.org/10.1007/s00382-020-05265-8.

    • Search Google Scholar
    • Export Citation
  • Luo, J.-J., G. Liu, H. Hendon, O. Alves, and T. Yamagata, 2017: Inter-basin sources for two-year predictability of the multi-year La Niña event in 2010–2012. Sci. Rep., 7, 2276, https://doi.org/10.1038/s41598-017-01479-9.

    • Search Google Scholar
    • Export Citation
  • Luseno, W. K., J. G. McPeak, C. B. Barrett, P. D. Little, and G. Gebru, 2003: Assessing the value of climate forecast information for pastoralists: Evidence from southern Ethiopia and northern Kenya. World Dev., 31, 14771494, https://doi.org/10.1016/S0305-750X(03)00113-X.

    • Search Google Scholar
    • Export Citation
  • Lyon, B., and D. G. DeWitt, 2012: A recent and abrupt decline in the East African long rains. Geophys. Res. Lett., 39, L02702, https://doi.org/10.1029/2011GL050337.

    • Search Google Scholar
    • Export Citation
  • Maxwell, D., and N. Majid, 2016: Famine in Somalia: Competing Imperatives, Collective Failures, 2011–2012. Oxford University Press, 269 pp.

  • Maxwell, D., and P. Hailey, 2020: The politics of information and analysis in famines and extreme emergencies synthesis of findings from six case studies. Feinstein International Center Tech. Rep., 50 pp., https://fic.tufts.edu/wp-content/uploads/PIA-Synthesis-Report_May-13.pdf.

  • Nicholson, S. E., 2017: Climate and climatic variability of rainfall over eastern Africa. Rev. Geophys., 55, 590635, https://doi.org/10.1002/2016RG000544.

    • Search Google Scholar
    • Export Citation
  • Okoola, R. E., 1999a: A diagnostic study of the eastern Africa monsoon circulation during the Northern Hemisphere spring season. Int. J. Climatol., 19, 143168, https://doi.org/10.1002/(SICI)1097-0088(199902)19:2%3C143::AID-JOC342%3E3.0.CO;2-U.

    • Search Google Scholar
    • Export Citation
  • Okoola, R. E., 1999b: Midtropospheric circulation patterns associated with extreme dry and wet episodes over equatorial eastern Africa during the Northern Hemisphere spring. J. Appl. Meteor. Climatol., 38, 11611169, https://doi.org/10.1175/1520-0450(1999)038<1161:MCPAWE>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Okumura, Y. M., and C. Deser, 2010: Asymmetry in the duration of El Niño and La Niña. J. Climate, 23, 58265843, https://doi.org/10.1175/2010JCLI3592.1.

    • Search Google Scholar
    • Export Citation
  • Okumura, Y. M., P. DiNezio, and C. Deser, 2017: Evolving impacts of multiyear La Niña events on atmospheric circulation and U.S. drought. Geophys. Res. Lett., 44, 11 61411 623, https://doi.org/10.1002/2017GL075034.

    • Search Google Scholar
    • Export Citation
  • Rao, M. P., E. R. Cook, B. I. Cook, K. J. Anchukaitis, R. D. D’Arrigo, P. J. Krusic, and A. N. LeGrande, 2019: A double bootstrap approach to superposed epoch analysis to evaluate response uncertainty. Dendrochronologia, 55, 119124, https://doi.org/10.1016/j.dendro.2019.05.001.

    • Search Google Scholar
    • Export Citation
  • Rasmusson, E. M., and T. H. Carpenter, 1982: Variations in tropical sea surface temperature and surface wind fields associated with the Southern Oscillation/El Niño. Mon. Wea. Rev., 110, 354384, https://doi.org/10.1175/1520-0493(1982)110<0354:VITSST>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Rayner, N. A., D. E. Parker, E. B. Horton, C. K. Folland, L. V. Alexander, D. P. Rowell, E. C. Kent, and A. Kaplan, 2003: Global analyses of sea surface temperature, sea ice, and night marine air temperature since the late nineteenth century. J. Geophys. Res., 108, 4407, https://doi.org/10.1029/2002JD002670.

    • Search Google Scholar
    • Export Citation
  • Rufino, M. C., and Coauthors, 2013: Transitions in agro-pastoralist systems of East Africa: Impacts on food security and poverty. Agric. Ecosyst. Environ., 179, 215230, https://doi.org/10.1016/j.agee.2013.08.019.

    • Search Google Scholar
    • Export Citation
  • Saji, N. H., and T. Yamagata, 2003: Possible impacts of Indian Ocean dipole mode events on global climate. Climate Res., 25, 151169, https://doi.org/10.3354/cr025151.

    • Search Google Scholar
    • Export Citation
  • Schneider, U., A. Becker, P. Finger, A. Meyer-Christoffer, and M. Ziese, 2018: GPCC full data monthly product version 2018 at 0.5°: Monthly land-surface precipitation from rain-gauges built on GTS-based and historical data. Global Precipitation Climatology Centre, accessed 1 June 2021, https://opendata.dwd.de/climate_environment/GPCC/html/fulldata-monthly_v2018_doi_download.html.

  • Seager, R., M. Cane, N. Henderson, D.-E. Lee, R. Abernathey, and H. Zhang, 2019: Strengthening tropical Pacific zonal sea surface temperature gradient consistent with rising greenhouse gases. Nat. Climate Change, 9, 517522, https://doi.org/10.1038/s41558-019-0505-x.

    • Search Google Scholar
    • Export Citation
  • Shukla, S., C. Funk, and A. Hoell, 2014: Using constructed analogs to improve the skill of National Multi-Model Ensemble March–April–May precipitation forecasts in equatorial East Africa. Environ. Res. Lett., 9, 094009, https://doi.org/10.1088/1748-9326/9/9/094009.

    • Search Google Scholar
    • Export Citation
  • Shukla, S., J. Roberts, A. Hoell, C. C. Funk, F. Robertson, and B. Kirtman, 2019: Assessing North American multimodel ensemble (NMME) seasonal forecast skill to assist in the early warning of anomalous hydrometeorological events over East Africa. Climate Dyn., 53, 74117427, https://doi.org/10.1007/s00382-016-3296-z.

    • Search Google Scholar
    • Export Citation
  • Shukla, S., G. Husak, W. Turner, F. Davenport, C. Funk, L. Harrison, and N. Krell, 2021: A slow rainy season onset is a reliable harbinger of drought in most food insecure regions in Sub-Saharan Africa. PLOS ONE, 16, e0242883, https://doi.org/10.1371/journal.pone.0242883.

    • Search Google Scholar
    • Export Citation
  • Tierney, J. E., J. E. Smerdon, K. J. Anchukaitis, and R. Seager, 2013: Multidecadal variability in East African hydroclimate controlled by the Indian Ocean. Nature, 493, 389392, https://doi.org/10.1038/nature11785.

    • Search Google Scholar
    • Export Citation
  • Tokinaga, H., I. Richter, and Y. Kosaka, 2019: ENSO influence on the Atlantic Niño, revisited: Multi-year versus single-year ENSO events. J. Climate, 32, 45854600, https://doi.org/10.1175/JCLI-D-18-0683.1.

    • Search Google Scholar
    • Export Citation
  • Ummenhofer, C. C., M. Kulüke, and J. E. Tierney, 2018: Extremes in East African hydroclimate and links to Indo-Pacific variability on interannual to decadal timescales. Climate Dyn., 50, 29712991, https://doi.org/10.1007/s00382-017-3786-7.

    • Search Google Scholar
    • Export Citation
  • Williams, A. P., and C. Funk, 2011: A westward extension of the warm pool leads to a westward extension of the Walker circulation, drying eastern Africa. Climate Dyn., 37, 24172435, https://doi.org/10.1007/s00382-010-0984-y.

    • Search Google Scholar
    • Export Citation
  • Wu, X., Y. M. Okumura, and P. N. DiNezio, 2019: What controls the duration of El Niño and La Niña events? J. Climate, 32, 59415965, https://doi.org/10.1175/JCLI-D-18-0681.1.

    • Search Google Scholar
    • Export Citation
  • Wu, X., Y. M. Okumura, C. Deser, and P. N. DiNezio, 2021: Two-year dynamical predictions of ENSO event duration during 1954–2015. J. Climate, 34, 40694087, https://doi.org/10.1175/JCLI-D-20-0619.1.

    • Search Google Scholar
    • Export Citation
  • Yang, W., R. Seager, M. A. Cane, and B. Lyon, 2014: The East African long rains in observations and models. J. Climate, 27, 71857202, https://doi.org/10.1175/JCLI-D-13-00447.1.

    • Search Google Scholar
    • Export Citation
  • Yang, W., R. Seager, M. A. Cane, and B. Lyon, 2015: The annual cycle of East African precipitation. J. Climate, 28, 23852404, https://doi.org/10.1175/JCLI-D-14-00484.1.

    • Search Google Scholar
    • Export Citation
  • Zebiak, S. E., and M. A. Cane, 1987: A model El Niño–Southern Oscillation. Mon. Wea. Rev., 115, 22622278, https://doi.org/10.1175/1520-0493(1987)115<2262:AMENO>2.0.CO;2.

    • Search Google Scholar
    • Export Citation

Supplementary Materials

Save
  • Anderson, W., and Coauthors, 2021: Violent conflict exacerbated drought-related food insecurity between 2009 and 2019 in sub-Saharan Africa. Nat. Food, 2, 603615, https://doi.org/10.1038/s43016-021-00327-4.

    • Search Google Scholar
    • Export Citation
  • Barnston, A. G., M. K. Tippett, M. L. L’Heureux, S. Li, and D. G. DeWitt, 2012: Skill of real-time seasonal ENSO model predictions during 2002–11: Is our capability increasing? Bull. Amer. Meteor. Soc., 93, 631651, https://doi.org/10.1175/BAMS-D-11-00111.1.

    • Search Google Scholar
    • Export Citation
  • Barnston, A. G., M. K. Tippett, M. Ranganathan, and M. L. L’Heureux, 2019: Deterministic skill of ENSO predictions from the North American multimodel ensemble. Climate Dyn., 53, 72157234, https://doi.org/10.1007/s00382-017-3603-3.

    • Search Google Scholar
    • Export Citation
  • Behera, S. K., J.-J. Luo, S. Masson, P. Delecluse, S. Gualdi, A. Navarra, and T. Yamagata, 2005: Paramount impact of the Indian Ocean dipole on the East African short rains: A CGCM study. J. Climate, 18, 45144530, https://doi.org/10.1175/JCLI3541.1.

    • Search Google Scholar
    • Export Citation
  • Blau, M. T., and K.-J. Ha, 2020: The Indian Ocean dipole and its impact on East African short rains in two CMIP5 historical scenarios with and without anthropogenic influence. J. Geophys. Res. Atmos., 125, e2020JD033121, https://doi.org/10.1029/2020JD033121.

    • Search Google Scholar
    • Export Citation
  • Camberlin, P., and R. Okoola, 2003: The onset and cessation of the “long rains” in eastern Africa and their interannual variability. Theor. Appl. Climatol., 75, 4354, https://doi.org/10.1007/s00704-002-0721-5.

    • Search Google Scholar
    • Export Citation
  • Clement, A. C., R. Seager, M. A. Cane, and S. E. Zebiak, 1996: An ocean dynamical thermostat. J. Climate, 9, 21902196, https://doi.org/10.1175/1520-0442(1996)009<2190:AODT>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • DiNezio, P. N., and C. Deser, 2014: Nonlinear controls on the persistence of La Niña. J. Climate, 27, 73357355, https://doi.org/10.1175/JCLI-D-14-00033.1.

    • Search Google Scholar
    • Export Citation
  • DiNezio, P. N., C. Deser, Y. Okumura, and A. Karspeck, 2017a: Predictability of 2-year La Niña events in a coupled general circulation model. Climate Dyn., 49, 42374261, https://doi.org/10.1007/s00382-017-3575-3.

    • Search Google Scholar
    • Export Citation
  • DiNezio, P. N., and Coauthors, 2017b: A 2 year forecast for a 60–80% chance of La Niña in 2017–2018. Geophys. Res. Lett., 44, 11 62411 635, https://doi.org/10.1002/2017GL074904.

    • Search Google Scholar
    • Export Citation
  • Dutra, E., L. Magnusson, F. Wetterhall, H. L. Cloke, G. Balsamo, S. Boussetta, and F. Pappenberger, 2013: The 2010–2011 drought in the Horn of Africa in ECMWF reanalysis and seasonal forecast products. Int. J. Climatol., 33, 17201729, https://doi.org/10.1002/joc.3545.

    • Search Google Scholar
    • Export Citation
  • FEWS NET, 2021: The eastern Horn of Africa faces an exceptional prolonged and persistent agro-pastoral drought sequence. ICPAC/FEWS NET/FAO GIEWS/WFP/JRC Tech. Rep., 8 pp., https://mars.jrc.ec.europa.eu/asap/files/special_focus_2021_11.pdf.

  • Funk, C., and A. Hoell, 2015: The leading mode of observed and CMIP5 ENSO-residual sea surface temperatures and associated changes in Indo-Pacific climate. J. Climate, 28, 43094329, https://doi.org/10.1175/JCLI-D-14-00334.1.

    • Search Google Scholar
    • Export Citation
  • Funk, C., A. Hoell, S. Shukla, I. Blade, B. Liebmann, J. B. Roberts, F. R. Robertson, and G. Husak, 2014: Predicting East African spring droughts using Pacific and Indian Ocean sea surface temperature indices. Hydrol. Earth Syst. Sci., 18, 49654978, https://doi.org/10.5194/hess-18-4965-2014.

    • Search Google Scholar
    • Export Citation
  • Funk, C., and Coauthors, 2018: Examining the role of unusually warm Indo-Pacific sea-surface temperatures in recent African droughts. Quart. J. Roy. Meteor. Soc., 144, 360383, https://doi.org/10.1002/qj.3266.

    • Search Google Scholar
    • Export Citation
  • Funk, C., and Coauthors, 2019: Examining the potential contributions of extreme “Western V” sea surface temperatures to the 2017 March–June East African drought. Bull. Amer. Meteor. Soc., 100 (1), S55S60, https://doi.org/10.1175/BAMS-D-18-0108.1.

    • Search Google Scholar
    • Export Citation
  • Ghiggi, G., V. Humphrey, S. I. Seneviratne, and L. Gudmundsson, 2019: GRUN: An observation-based global gridded runoff dataset from 1902 to 2014. Earth Syst. Sci. Data, 11, 16551674, https://doi.org/10.5194/essd-11-1655-2019.

    • Search Google Scholar
    • Export Citation
  • Gleixner, S., T. Demissie, and G. T. Diro, 2020: Did ERA5 improve temperature and precipitation reanalysis over East Africa? Atmosphere, 11, 996, https://doi.org/10.3390/atmos11090996.

    • Search Google Scholar
    • Export Citation
  • Goddard, L., and N. E. Graham, 1999: Importance of the Indian Ocean for simulating rainfall anomalies over eastern and southern Africa. J. Geophys. Res., 104, 19 09919 116, https://doi.org/10.1029/1999JD900326.

    • Search Google Scholar
    • Export Citation
  • Ham, Y.-G., J.-H. Kim, and J.-J. Luo, 2019: Deep learning for multi-year ENSO forecasts. Nature, 573, 568572, https://doi.org/10.1038/s41586-019-1559-7.

    • Search Google Scholar
    • Export Citation
  • Harris, I., T. J. Osborn, P. Jones, and D. Lister, 2020: Version 4 of the CRU TS monthly high-resolution gridded multivariate climate dataset. Sci. Data, 7, 109, https://doi.org/10.1038/s41597-020-0453-3.

    • Search Google Scholar
    • Export Citation
  • Hastenrath, S., D. Polzin, and C. Mutai, 2011: Circulation mechanisms of Kenya rainfall anomalies. J. Climate, 24, 404412, https://doi.org/10.1175/2010JCLI3599.1.

    • Search Google Scholar
    • Export Citation
  • Hersbach, H., and Coauthors, 2020: The ERA5 global reanalysis. Quart. J. Roy. Meteor. Soc., 146, 19992049, https://doi.org/10.1002/qj.3803.

    • Search Google Scholar
    • Export Citation
  • Hoell, A., and C. Funk, 2013: The ENSO-related west Pacific sea surface temperature gradient. J. Climate, 26, 95459562, https://doi.org/10.1175/JCLI-D-12-00344.1.

    • Search Google Scholar
    • Export Citation
  • Hoell, A., and C. Funk, 2014: Indo-Pacific sea surface temperature influences on failed consecutive rainy seasons over eastern Africa. Climate Dyn., 43, 16451660, https://doi.org/10.1007/s00382-013-1991-6.

    • Search Google Scholar
    • Export Citation
  • Hoell, A., C. Funk, and M. Barlow, 2014: The regional forcing of Northern Hemisphere drought during recent warm tropical west Pacific Ocean La Niña events. Climate Dyn., 42, 32893311, https://doi.org/10.1007/s00382-013-1799-4.

    • Search Google Scholar
    • Export Citation
  • Jong, B.-T., M. Ting, R. Seager, and W. B. Anderson, 2020: ENSO teleconnections and impacts on U.S summertime temperature during a multiyear La Niña life cycle. J. Climate, 33, 60096024, https://doi.org/10.1175/JCLI-D-19-0701.1.

    • Search Google Scholar
    • Export Citation
  • Lenssen, N. J., L. Goddard, and S. Mason, 2020: Seasonal forecast skill of ENSO teleconnection maps. Wea. Forecasting, 35, 23872406, https://doi.org/10.1175/WAF-D-19-0235.1.

    • Search Google Scholar
    • Export Citation
  • Liebmann, B., and Coauthors, 2014: Understanding recent eastern Horn of Africa rainfall variability and change. J. Climate, 27, 86308645, https://doi.org/10.1175/JCLI-D-13-00714.1.

    • Search Google Scholar
    • Export Citation
  • Liebmann, B., and Coauthors, 2017: Climatology and interannual variability of boreal spring wet season precipitation in the eastern Horn of Africa and implications for its recent decline. J. Climate, 30, 38673886, https://doi.org/10.1175/JCLI-D-16-0452.1.

    • Search Google Scholar
    • Export Citation
  • Liu, W., K. H. Cook, and E. K. Vizy, 2020: Influence of Indian Ocean SST regionality on the East African short rains. Climate Dyn., 54, 49915011, https://doi.org/10.1007/s00382-020-05265-8.

    • Search Google Scholar
    • Export Citation
  • Luo, J.-J., G. Liu, H. Hendon, O. Alves, and T. Yamagata, 2017: Inter-basin sources for two-year predictability of the multi-year La Niña event in 2010–2012. Sci. Rep., 7, 2276, https://doi.org/10.1038/s41598-017-01479-9.

    • Search Google Scholar
    • Export Citation
  • Luseno, W. K., J. G. McPeak, C. B. Barrett, P. D. Little, and G. Gebru, 2003: Assessing the value of climate forecast information for pastoralists: Evidence from southern Ethiopia and northern Kenya. World Dev., 31, 14771494, https://doi.org/10.1016/S0305-750X(03)00113-X.

    • Search Google Scholar
    • Export Citation
  • Lyon, B., and D. G. DeWitt, 2012: A recent and abrupt decline in the East African long rains. Geophys. Res. Lett., 39, L02702, https://doi.org/10.1029/2011GL050337.

    • Search Google Scholar
    • Export Citation
  • Maxwell, D., and N. Majid, 2016: Famine in Somalia: Competing Imperatives, Collective Failures, 2011–2012. Oxford University Press, 269 pp.

  • Maxwell, D., and P. Hailey, 2020: The politics of information and analysis in famines and extreme emergencies synthesis of findings from six case studies. Feinstein International Center Tech. Rep., 50 pp., https://fic.tufts.edu/wp-content/uploads/PIA-Synthesis-Report_May-13.pdf.

  • Nicholson, S. E., 2017: Climate and climatic variability of rainfall over eastern Africa. Rev. Geophys., 55, 590635, https://doi.org/10.1002/2016RG000544.

    • Search Google Scholar
    • Export Citation
  • Okoola, R. E., 1999a: A diagnostic study of the eastern Africa monsoon circulation during the Northern Hemisphere spring season. Int. J. Climatol., 19, 143168, https://doi.org/10.1002/(SICI)1097-0088(199902)19:2%3C143::AID-JOC342%3E3.0.CO;2-U.

    • Search Google Scholar
    • Export Citation
  • Okoola, R. E., 1999b: Midtropospheric circulation patterns associated with extreme dry and wet episodes over equatorial eastern Africa during the Northern Hemisphere spring. J. Appl. Meteor. Climatol., 38, 11611169, https://doi.org/10.1175/1520-0450(1999)038<1161:MCPAWE>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Okumura, Y. M., and C. Deser, 2010: Asymmetry in the duration of El Niño and La Niña. J. Climate, 23, 58265843, https://doi.org/10.1175/2010JCLI3592.1.

    • Search Google Scholar
    • Export Citation
  • Okumura, Y. M., P. DiNezio, and C. Deser, 2017: Evolving impacts of multiyear La Niña events on atmospheric circulation and U.S. drought. Geophys. Res. Lett., 44, 11 61411 623, https://doi.org/10.1002/2017GL075034.

    • Search Google Scholar
    • Export Citation
  • Rao, M. P., E. R. Cook, B. I. Cook, K. J. Anchukaitis, R. D. D’Arrigo, P. J. Krusic, and A. N. LeGrande, 2019: A double bootstrap approach to superposed epoch analysis to evaluate response uncertainty. Dendrochronologia, 55, 119124, https://doi.org/10.1016/j.dendro.2019.05.001.

    • Search Google Scholar
    • Export Citation
  • Rasmusson, E. M., and T. H. Carpenter, 1982: Variations in tropical sea surface temperature and surface wind fields associated with the Southern Oscillation/El Niño. Mon. Wea. Rev., 110, 354384, https://doi.org/10.1175/1520-0493(1982)110<0354:VITSST>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Rayner, N. A., D. E. Parker, E. B. Horton, C. K. Folland, L. V. Alexander, D. P. Rowell, E. C. Kent, and A. Kaplan, 2003: Global analyses of sea surface temperature, sea ice, and night marine air temperature since the late nineteenth century. J. Geophys. Res., 108, 4407, https://doi.org/10.1029/2002JD002670.

    • Search Google Scholar
    • Export Citation
  • Rufino, M. C., and Coauthors, 2013: Transitions in agro-pastoralist systems of East Africa: Impacts on food security and poverty. Agric. Ecosyst. Environ., 179, 215230, https://doi.org/10.1016/j.agee.2013.08.019.

    • Search Google Scholar
    • Export Citation
  • Saji, N. H., and T. Yamagata, 2003: Possible impacts of Indian Ocean dipole mode events on global climate. Climate Res., 25, 151169, https://doi.org/10.3354/cr025151.

    • Search Google Scholar
    • Export Citation
  • Schneider, U., A. Becker, P. Finger, A. Meyer-Christoffer, and M. Ziese, 2018: GPCC full data monthly product version 2018 at 0.5°: Monthly land-surface precipitation from rain-gauges built on GTS-based and historical data. Global Precipitation Climatology Centre, accessed 1 June 2021, https://opendata.dwd.de/climate_environment/GPCC/html/fulldata-monthly_v2018_doi_download.html.

  • Seager, R., M. Cane, N. Henderson, D.-E. Lee, R. Abernathey, and H. Zhang, 2019: Strengthening tropical Pacific zonal sea surface temperature gradient consistent with rising greenhouse gases. Nat. Climate Change, 9, 517522, https://doi.org/10.1038/s41558-019-0505-x.

    • Search Google Scholar
    • Export Citation
  • Shukla, S., C. Funk, and A. Hoell, 2014: Using constructed analogs to improve the skill of National Multi-Model Ensemble March–April–May precipitation forecasts in equatorial East Africa. Environ. Res. Lett., 9, 094009, https://doi.org/10.1088/1748-9326/9/9/094009.

    • Search Google Scholar
    • Export Citation
  • Shukla, S., J. Roberts, A. Hoell, C. C. Funk, F. Robertson, and B. Kirtman, 2019: Assessing North American multimodel ensemble (NMME) seasonal forecast skill to assist in the early warning of anomalous hydrometeorological events over East Africa. Climate Dyn., 53, 74117427, https://doi.org/10.1007/s00382-016-3296-z.

    • Search Google Scholar
    • Export Citation
  • Shukla, S., G. Husak, W. Turner, F. Davenport, C. Funk, L. Harrison, and N. Krell, 2021: A slow rainy season onset is a reliable harbinger of drought in most food insecure regions in Sub-Saharan Africa. PLOS ONE, 16, e0242883, https://doi.org/10.1371/journal.pone.0242883.

    • Search Google Scholar
    • Export Citation
  • Tierney, J. E., J. E. Smerdon, K. J. Anchukaitis, and R. Seager, 2013: Multidecadal variability in East African hydroclimate controlled by the Indian Ocean. Nature, 493, 389392, https://doi.org/10.1038/nature11785.

    • Search Google Scholar
    • Export Citation
  • Tokinaga, H., I. Richter, and Y. Kosaka, 2019: ENSO influence on the Atlantic Niño, revisited: Multi-year versus single-year ENSO events. J. Climate, 32, 45854600, https://doi.org/10.1175/JCLI-D-18-0683.1.

    • Search Google Scholar
    • Export Citation
  • Ummenhofer, C. C., M. Kulüke, and J. E. Tierney, 2018: Extremes in East African hydroclimate and links to Indo-Pacific variability on interannual to decadal timescales. Climate Dyn., 50, 29712991, https://doi.org/10.1007/s00382-017-3786-7.

    • Search Google Scholar
    • Export Citation
  • Williams, A. P., and C. Funk, 2011: A westward extension of the warm pool leads to a westward extension of the Walker circulation, drying eastern Africa. Climate Dyn., 37, 24172435, https://doi.org/10.1007/s00382-010-0984-y.

    • Search Google Scholar
    • Export Citation
  • Wu, X., Y. M. Okumura, and P. N. DiNezio, 2019: What controls the duration of El Niño and La Niña events? J. Climate, 32, 59415965, https://doi.org/10.1175/JCLI-D-18-0681.1.

    • Search Google Scholar
    • Export Citation
  • Wu, X., Y. M. Okumura, C. Deser, and P. N. DiNezio, 2021: Two-year dynamical predictions of ENSO event duration during 1954–2015. J. Climate, 34, 40694087, https://doi.org/10.1175/JCLI-D-20-0619.1.

    • Search Google Scholar
    • Export Citation
  • Yang, W., R. Seager, M. A. Cane, and B. Lyon, 2014: The East African long rains in observations and models. J. Climate, 27, 71857202, https://doi.org/10.1175/JCLI-D-13-00447.1.

    • Search Google Scholar
    • Export Citation
  • Yang, W., R. Seager, M. A. Cane, and B. Lyon, 2015: The annual cycle of East African precipitation. J. Climate, 28, 23852404, https://doi.org/10.1175/JCLI-D-14-00484.1.

    • Search Google Scholar
    • Export Citation
  • Zebiak, S. E., and M. A. Cane, 1987: A model El Niño–Southern Oscillation. Mon. Wea. Rev., 115, 22622278, https://doi.org/10.1175/1520-0493(1987)115<2262:AMENO>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Fig. 1.

    (a),(b) Sea surface temperature anomalies in the Niño-3.4 region, (c),(d) precipitation anomalies in the Horn of Africa, and (e),(f) runoff anomalies in the Horn of Africa during (left) first-year and second-year events of multiyear La Niñas and (right) single-year La Niñas. Individual events shown with dotted lines and composite means shown with solid lines. The Horn of Africa is defined as the region between 38° and 53°E and between 5°S and 8°N.

  • Fig. 2.

    Uncertainty around composite means from Fig. 1 estimated using a double-bootstrap superposed epoch analysis. Shading shows the 5th, 10th, 20th, and 80th, 90th, and 95th percentiles of the composite mean, while the solid lines are composite means, as in Fig. 1 for (a),(b) sea surface temperature anomalies in the Niño-3.4 region, (c),(d) precipitation anomalies in the Horn of Africa, and (e),(f) runoff anomalies in the Horn of Africa during (left) first-year and second-year events of multiyear La Niñas and (right) single-year La Niñas. Dotted lines represent bootstrapped uncertainty estimates as the 5th and 95th (black), 10th and 90th (gray), or 20th and 80th (light gray) percentiles of composite means calculated based on years drawn at random from the full record (see methods).

  • Fig. 3.

    Average (a),(c) precipitation and (b),(d) runoff z scores in the Horn of Africa during single-year La Niñas during the OND short rains and the subsequent MAM long rains. Gray regions indicate climatologically dry areas, defined as receiving less than 30% of annual precipitation during the plotted months.

  • Fig. 4.

    Average (left) precipitation and (right) runoff z scores in the Horn of Africa during multiyear La Niñas (a)–(d) during the OND short rains of first-year events and the subsequent the MAM long rains during first-year events, as well as (e)–(h) during the OND short rains of second-year events and the subsequent the MAM long rains during second-year events. Gray regions indicate climatologically dry areas, defined as receiving less than 30% of annual precipitation during the plotted months.

  • Fig. 5.

    Sea surface temperature anomalies during the (left) short rains and (right) long rains of (a),(b) single-year La Niñas as well as during the (c),(d) first year and (e),(f) second year of multiyear La Niñas. Hatching indicates SST anomalies that are not significant at the p < 0.9 level.

  • Fig. 6.

    Atmospheric circulation (a) climatology and (b) anomalies averaged from 5°S to 5°N during single-year La Niñas, as well as during the (c) first year and (d) second year of multiyear La Niñas.

  • Fig. 7.

    Vertically integrated moisture transport—defined as moisture transport integrated from the surface of Earth to the top of the atmosphere—(a),(b) climatology and (c),(d) anomalies during single-year La Niñas as well as during the (e),(f) first year and (g),(h) second year of multiyear La Niñas. Shading in each plot shows the magnitude of transport. In (c)–(h), positive values indicate transport that enhances the mean-state transport while negative values indicate moisture transport that is in the opposite sense of the mean-state transport. (left) All Africa domain; (right) the same quantities over East Africa.