< Back to previous page
Publication
Patterns and mechanisms of climate extremes in East Africa. Final Report.
Book - Report
Subtitle:PAMEXEA. Contract - BR/121/A2. BRAIN-be - Belgian Research Action through Interdisciplinary Networks
Context. Catastrophic droughts afflicting the Horn of Africa in recent years highlight the heavy impact of a variable climate on the region’s vulnerable populations and socio-economic systems; and the enormous challenge to develop a sustainable agricultural economy in a future of climate change, growing demographic pressure and naturally scarce water resources. Most worrying is the uncertain impact of climate change on regional freshwater resources, due to incomplete understanding of the effects of a warming atmosphere on the region’s hydro-climate. Global climate-model (GCM) simulations underpinning current IPCC prognoses for East Africa have insufficient spatial downscaling to guide the water-resource planning needed by developing economies to prepare for the future.
Objectives. Applying the powerful ‘the past is the key to the (present and) future’ approach, PAMEXEA aimed to improve understanding of climate trends and extremes in East Africa at multiple time scales as foundation for appropriate water-resource management in both dry and wetter regions of East Africa. Specifically, we aimed to find out how processes of natural climate variability at inter-annual, decadal and centennial time scales may interact with anthropogenic climate forcing to create future trends in rainfall and drought. In addition we aimed to place the current magnitude of human demands on East Africa’s natural resources in a proper long-term context by reconstructing the history of anthropogenic impact on the region’s aquatic and terrestrial environments. PAMEXEA consisted of distinct work packages (WPs) focusing on i) the quality screening and compilation of existing climate-indicator (‘climate-proxy’) data, ii) new hydro-climate (i.e., moisture-balance) reconstructions from carefully selected African lake-sediment records, iii) long-term hydrological and ecological monitoring of lake systems targeted for sediment-based climate reconstruction, iv) hydrological modelling of those lake systems to assess their sensitivity to hydro-climate change, and v) testing the ability of existing climate models to replicate the temporal and spatial patterns of past hydro-climate variability.
Originally focused on equatorial East Africa and on the last 2000 years, exquisite lake-sediment records obtained through international collaborations allowed expansion of the project’s scope to include also the Sahara desert of North Africa and the major, century-scale climate extremes which have impacted both East and North Africa throughout the Holocene period (the last 11,650 years). These century-scale droughts epitomise how ‘abrupt’ climate change and climate extremes have impacted on pre-industrial societies in arid and semi-arid tropical regions worldwide, and may do so again in the future.
Results and Conclusions. Field- and lab-based PAMEXEA activities produced new high-resolution hydro-climate reconstructions from five lakes: Bogoria in central Kenya (the last 1300 years), Chala in south-eastern Kenya (2200 years), Nasiki Engida in south-western Kenya (2800 years), Ounianga Serir in northern Chad (4200 years) and Rutundu on Mt. Kenya (18,000 years). We also upgraded the existing reconstruction from Lake Naivasha in central Kenya (1650 years) using new palaeo-hydrological proxies and age-modelling. Using our hydro-climate reconstructions from lakes Naivasha, Bogoria and Chala as reference frame, we also reconstructed the history of ancient and historical human impact on the landscape of central and south-eastern Kenya in relation to the magnitude of natural aquatic and terrestrial ecosystem dynamics. The principal new insights of this part of the PAMEXEA project are that I) the multi-decadal events of severe drought which occurred ~8200, ~4200 and ~2900 years ago and were felt across tropical Africa were pulsed negative rainfall anomalies superimposed upon a long-term trend of gradual Holocene drying, not (as widely believed) abrupt transitions between two contrasting climate states before and after the events; II) when viewed over the entire Holocene period, driest climatic conditions occurred ~2900 years ago in East Africa and ~4200 years ago in North Africa; III) human impact on the landscape (i.e., on terrestrial vegetation) of equatorial East Africa becomes noticeable from between ~900 and ~250 years ago with early appearances mostly limited to areas with above-average population density; and IV) human activity has become the principal driver of terrestrial and aquatic ecosystem dynamics only within the last century, most often in the last ~50-60 years.
Integrating our new reconstructions for East Africa with published paleoclimate records produced a synthesis of hydro-climate variability across the African continent in recent millennia. To assess the magnitude of rainfall variation at this time scale, we fine-tuned an existing water-balance model for Lake Naivasha and applied it to the published lake-level reconstruction. The main conclusions of these exercises are that V) natural, decade-to-century scale climate variability during the last millennium has been rather uniform across the ‘Horn of Africa’ region of East Africa where all rainfall is sourced in the Indian Ocean; and VI) the full amplitude of lake-level fluctuation recorded over the past 1100 years can be explained by relatively modest rainfall variability of ~10% at decadal time scales.
PAMEXEA conducted the last five years (2013-2017) of an 11-year monitoring program on Lake Chala, thereby enabling the calibration of local phytoplankton succession against seasonal and inter-annual weather patterns for use of phytoplankton fossils (diatoms, in particular) preserved in the sediments as climate proxy. In collaboration with partners at Utrecht University this calibration was expanded to selected bacterial biomarkers to promote also their use as climate proxies. The principal new insights of this part of the work programme are that VII) some diatoms dominate dry-season algal productivity in Lake Chala during La Niña-type years with weak ‘long’ rains and a long dry season, whereas others are dominant during years with less wind, and thus more shallow water-column mixing; and VIII) the so-called BIT index of bacterial lipids in Lake Chala sediments, an important proxy for past rainfall variation, is controlled by the influence of rainfall on water-column stratification and oxygenation.
Improvements in long-term prediction of climate change and weather extremes in East Africa depend on climate models being capable of capturing the principal large-scale climate-dynamical processes, the mechanisms of tropical climate forcing at inter-annual to centennial time scales, and all region-specific feedbacks and land-ocean-atmosphere interactions. The principal new insights derived from PAMEXEA effort in this field are that IX) state-of-the-art global climate models (GCMs) simulate fairly well the unimodal annual cycle of modern-day precipitation in south-eastern Africa, while the bimodal rainfall cycle in eastern equatorial Africa is less well captured; X) climate-model simulations over the last 1000 years display very different temporal patterns than proxy-based reconstructions, and there is no common signal among the model time series prior to AD 1850, suggesting that simulated hydro-climate fluctuations are mostly driven by internal climate variability rather than by external (natural or anthropogenic) forcing; XI) the link between precipitation and Indian Ocean sea surface temperatures (SSTs) in pre-industrial time is stronger for equatorial East Africa than for south-eastern Africa, with East African rainfall being especially sensitive to an enhanced west-to-east Indian Ocean SST gradient; and XII) GCM simulations are unrealistic with respect to lake hydrology, and (even after adjustment) produce no shared lake-level signal nor do they resemble the proxy-based reconstruction.
Objectives. Applying the powerful ‘the past is the key to the (present and) future’ approach, PAMEXEA aimed to improve understanding of climate trends and extremes in East Africa at multiple time scales as foundation for appropriate water-resource management in both dry and wetter regions of East Africa. Specifically, we aimed to find out how processes of natural climate variability at inter-annual, decadal and centennial time scales may interact with anthropogenic climate forcing to create future trends in rainfall and drought. In addition we aimed to place the current magnitude of human demands on East Africa’s natural resources in a proper long-term context by reconstructing the history of anthropogenic impact on the region’s aquatic and terrestrial environments. PAMEXEA consisted of distinct work packages (WPs) focusing on i) the quality screening and compilation of existing climate-indicator (‘climate-proxy’) data, ii) new hydro-climate (i.e., moisture-balance) reconstructions from carefully selected African lake-sediment records, iii) long-term hydrological and ecological monitoring of lake systems targeted for sediment-based climate reconstruction, iv) hydrological modelling of those lake systems to assess their sensitivity to hydro-climate change, and v) testing the ability of existing climate models to replicate the temporal and spatial patterns of past hydro-climate variability.
Originally focused on equatorial East Africa and on the last 2000 years, exquisite lake-sediment records obtained through international collaborations allowed expansion of the project’s scope to include also the Sahara desert of North Africa and the major, century-scale climate extremes which have impacted both East and North Africa throughout the Holocene period (the last 11,650 years). These century-scale droughts epitomise how ‘abrupt’ climate change and climate extremes have impacted on pre-industrial societies in arid and semi-arid tropical regions worldwide, and may do so again in the future.
Results and Conclusions. Field- and lab-based PAMEXEA activities produced new high-resolution hydro-climate reconstructions from five lakes: Bogoria in central Kenya (the last 1300 years), Chala in south-eastern Kenya (2200 years), Nasiki Engida in south-western Kenya (2800 years), Ounianga Serir in northern Chad (4200 years) and Rutundu on Mt. Kenya (18,000 years). We also upgraded the existing reconstruction from Lake Naivasha in central Kenya (1650 years) using new palaeo-hydrological proxies and age-modelling. Using our hydro-climate reconstructions from lakes Naivasha, Bogoria and Chala as reference frame, we also reconstructed the history of ancient and historical human impact on the landscape of central and south-eastern Kenya in relation to the magnitude of natural aquatic and terrestrial ecosystem dynamics. The principal new insights of this part of the PAMEXEA project are that I) the multi-decadal events of severe drought which occurred ~8200, ~4200 and ~2900 years ago and were felt across tropical Africa were pulsed negative rainfall anomalies superimposed upon a long-term trend of gradual Holocene drying, not (as widely believed) abrupt transitions between two contrasting climate states before and after the events; II) when viewed over the entire Holocene period, driest climatic conditions occurred ~2900 years ago in East Africa and ~4200 years ago in North Africa; III) human impact on the landscape (i.e., on terrestrial vegetation) of equatorial East Africa becomes noticeable from between ~900 and ~250 years ago with early appearances mostly limited to areas with above-average population density; and IV) human activity has become the principal driver of terrestrial and aquatic ecosystem dynamics only within the last century, most often in the last ~50-60 years.
Integrating our new reconstructions for East Africa with published paleoclimate records produced a synthesis of hydro-climate variability across the African continent in recent millennia. To assess the magnitude of rainfall variation at this time scale, we fine-tuned an existing water-balance model for Lake Naivasha and applied it to the published lake-level reconstruction. The main conclusions of these exercises are that V) natural, decade-to-century scale climate variability during the last millennium has been rather uniform across the ‘Horn of Africa’ region of East Africa where all rainfall is sourced in the Indian Ocean; and VI) the full amplitude of lake-level fluctuation recorded over the past 1100 years can be explained by relatively modest rainfall variability of ~10% at decadal time scales.
PAMEXEA conducted the last five years (2013-2017) of an 11-year monitoring program on Lake Chala, thereby enabling the calibration of local phytoplankton succession against seasonal and inter-annual weather patterns for use of phytoplankton fossils (diatoms, in particular) preserved in the sediments as climate proxy. In collaboration with partners at Utrecht University this calibration was expanded to selected bacterial biomarkers to promote also their use as climate proxies. The principal new insights of this part of the work programme are that VII) some diatoms dominate dry-season algal productivity in Lake Chala during La Niña-type years with weak ‘long’ rains and a long dry season, whereas others are dominant during years with less wind, and thus more shallow water-column mixing; and VIII) the so-called BIT index of bacterial lipids in Lake Chala sediments, an important proxy for past rainfall variation, is controlled by the influence of rainfall on water-column stratification and oxygenation.
Improvements in long-term prediction of climate change and weather extremes in East Africa depend on climate models being capable of capturing the principal large-scale climate-dynamical processes, the mechanisms of tropical climate forcing at inter-annual to centennial time scales, and all region-specific feedbacks and land-ocean-atmosphere interactions. The principal new insights derived from PAMEXEA effort in this field are that IX) state-of-the-art global climate models (GCMs) simulate fairly well the unimodal annual cycle of modern-day precipitation in south-eastern Africa, while the bimodal rainfall cycle in eastern equatorial Africa is less well captured; X) climate-model simulations over the last 1000 years display very different temporal patterns than proxy-based reconstructions, and there is no common signal among the model time series prior to AD 1850, suggesting that simulated hydro-climate fluctuations are mostly driven by internal climate variability rather than by external (natural or anthropogenic) forcing; XI) the link between precipitation and Indian Ocean sea surface temperatures (SSTs) in pre-industrial time is stronger for equatorial East Africa than for south-eastern Africa, with East African rainfall being especially sensitive to an enhanced west-to-east Indian Ocean SST gradient; and XII) GCM simulations are unrealistic with respect to lake hydrology, and (even after adjustment) produce no shared lake-level signal nor do they resemble the proxy-based reconstruction.
Series: BRAIN-be - Belgian Research Action through Interdisciplinary Networks
Number of pages: 28
Publication year:2020
Accessibility:Open