"Abrupt onset and termination of the African Humid Period: rapid climate responses to gradual insolation forcing", P.D. deMenocal, J. Ortiz, T. Guilderson, J. Adkins, M. Sarnthein, L. Baker, M. Yarusinsky, Quaternary Science Reviews, 19:347-361 (2000).

Goal: document variability of African eolian dust in the marine record over last 25 k.y.

• Holocene African Humid Period: 9 - 6 ka BP; initiated 14.5 ka BP following hyperarid conditions of latest Pleistocene
• greening caused by intensification of African monsoon caused by 8% greater insolation than modern at 10-11 ka BP; alignment of boreal solstice with perihelion
• heating of North African land surface during boreal summer causes cyclonic flow from sea to land
• Prell and Kutzbach (1987): atmosphere only model; precipitation increases 5x over proportional increase of insolation
• only coupled atmosphere-ocean model can produce conditions required to create large North African lakes found in the geologic record
• positive feedback amplifications:
• increases in SST
• lower albedo; greater vegetation cover
• moisture availability
• ocean-atmosphere interactions are often quite non-linear; thresholds are crossed causing rapid, abrupt climate changes
• the shift from humid to arid conditions between 6 and 5 ka in response to declining summer insolation
• pollen levels show mesic plants disappearing
• sedentary lacustrine peoples replaced by nomadic pastoralists

• 50-100 yr sampling interval at ODP site 658C off Cap Blanc, Mauritania (20 degrees 45' N, 18 degrees 35' W in 2263 m water depth; 18 cm/ky SAR; coastal upwelling zone
• directly below axis of summer African dust plume; under easterly propogating African Easterly Jet (AEJ) w/ strongly turbulent surface winds; combination of turbulent suspension and vertical convection lifts dust to mid troposphere
• dust: silt-sized grains of subrounded quartz and feldspar and assoc. illite matrix; highly weathered source area

• sample interval 50 to 150 yrs; 2 cm spacing
• organic C: 0.5% - 1.5%
• terrigenous (detrital) fraction: determined by subtracting biogenic opal (measured with a spectrophotometric technique) from total silica, which is everything that isn't carbonate.
• benthic isotope record using cibs
• age control by C-14 on bulloides at 18 levels

• hiatus at 324-328 cm (14.69-17.21 ka); assoc. with step-like increases in several proxies; spans the first phase of deglaciation; may be artefact introduced by gas expansion
• 3 major compositional shifts

1. 324 cm (14.65 ka)
2. 260 cm (12.32 ka)
3. 125 cm (5.49 ka)

• 324-257 cm (14.65-5.49 ka): large increase in biogenic carbonate and opal (decr. in terrigenous sed); punctuated by brief incr. in terrigenous sed between 284-257 (13.38-12.32 ka) ... Younger Dryas

• 5.49 ka increase: 50% incr. in terrigenous sed flux
• 11.2-7.6 ka increase: 47 incr in terrigenous sed flux; 24% decr. in carbonate flux
• 14.8 ka decrease: 8% decr. in terrigenous sed; 59% incr. in carbonate

• transition completed in a few centuries
• 5.49 ka incr. took 4 centuries
• Younger Dryas arid period (12.3 ka) terminated in 2 centuries
• 14.8 ka transition complicated by hiatus, but agrees with other records of onset of the humid period

• Greenland ice cores record incr. in methane at 14.7 ka assoc. with growth of wetlands
• synchronous with end of cold, wet conditions in Europe and with Heinrich event 1
• Younger Dryas appears as brief return to aridity between 13.4 and 12.3 ka; this precedes the well-constrained YD by 800 yrs
• arid event at 8.2 ka is present in terrestrial records from Africa to Asia and corresponds with a brief cold event in the North Atlantic (coolest period of the Holocene) and spike in terrigenous input at Site 658C
• seems to have marked the beginning of the end of humid conditions in Africa; gradual decline in precip:evap balance until 5.5 ka, when a sharp increase in aridity occurs
• basin-to-basin variability of timing for shift to aridity between 6 and 5 ka
• no shift to lower methane levels in Greenland ice because high latitude wetlands expanded in late Holocene that were not present in earlier Holocene

• Africa entered and left the humid period at 470 W/m³ (4.2% above modern values); apparently a threshold value

• early modelling efforts suggested a linear relationship between insolation and African monsoon climate
• linear relationship hold for known range of late Pleistocene variations in insolation (-5% to 15%) ... Prell and Kutzbach (1987)
• simulated P:E budget could not produce humid climate though

• Kutzbach and Liu (1997): mid Holocene summer radiation increased SST 0.4 degrees C, which further increased African monsoonal precip by 25% over uncoupled model simulations
• monsoon intensified by incr. moist convection from warmer ocean surface
• rain penetrated deeper into Africa, but still did not match pluvial setting shown by paleoclimate records
• changes in African vegetation can increase monsoon precipitation as much as incr. in radiation alone
• Brostrom et al. (1998): changes in land-surface conditions plus coupled atmosphere-ocean model caused monsoon to be prolonged two months and penetrate 300 more km into the African interior

• CLIMBER(2) model trades resolution for multiple variable inclusion to capture transient climate changes
• CLIMBER2 reproduces the humid to arid transition near 5.5 ka by gradually dropping insolation values; similar in pattern to eolian dust supply data observed at Site 658C
• abruptness attributed to sensitivity of vegetation changes to precipitation fields; vegetation increases the efficiency of the insolation forcing

• two positive feedback processes contribute to the rapidity of the climate transitions in the Holocene:
• coupled vegetation-albedo feedback
• surface ocean temp-moisture transport feedback
• rectify sinusoidal insolation curve to make square-wave climate response signal