African humid period precipitation sustained by robust vegetation, soil and lake feedbacks
Deepak Chandan, Dick Peltier
Associated paper: Chandan & Peltier, 2020 GRL African humid period/Green Sahara
Evidence for green Sahara: •Vegetation reconstructions (Hoelzmann et al., 1998, Harrison & Bartlein 2012) •Palæo lake reconstructions (Lézine et al., 2011, Holmes & Hoelzmann, 2017) •Desiccated ancient river valleys (Neumann 1989, Kröpelin 2007) •Eolian deposits in sedimentary cores in the Atlantic (deMenocal et al., 2000) •Leaf wax deposits in sedimentary cores in the Atlantic (Tierney et al., 2017) •Cave paintings depicting lush landscapes and fauna (Barth 1857, Almásy 1934, di Lernia 2017) •Archæological findings supporting human habitation in presently uninhabited regions (Gabriel 1987, Hoelzmann et al., 2001, Kröpelin 2004, Serano et al., 2008, Dunne et al., 2012, Manning & Timpson, 2014)
Image: NASA African humid period/Green Sahara
Existing studies: •Feedbacks important
•Only minor enhancement to precipitation from orbital changes in the absence of feedbacks
•Feedbacks from land surface most important: Vegetation, lake and soil changes
•Extensively studied. BUT….
•Collective feedback from these feedbacks rarely studied
•Those that did, are very old now; revised assessment needed
•Understanding of mid-Holocene land surface has changed a lot since these studies
Image: NASA To simulate mid-Holocene Green Sahara we modify three Modeling the mid-Holocene Africa land surface properties, in addition to orbital parameters and GHGs: vegetation, soil and lakes. Our model is the We change soil composition and soil colour (albedo) UofT version of CCSM4. over northern Africa and the Arabian Peninsula. Dry soil albedo reduces from 18% to 9%.
Simulation Description
PIREF PI Control
C4 Grass C3 Grass Mid-Holocene Control C3 Arctic Grass MHREF BDS Boreal BDS Temperate BES Temperate BDT Boreal MH MHREF + Vegetation BDT Temperate V BDT Tropical BET Temperate BET Tropical MHV + Soil NDT Boreal MHVS NET Boreal NET Temperate BareSoil MHVL MHV + Lakes
mid-Holocene map of dominant mid-Holocene lakes MHVSL MHV + Soil + Lakes plant functional types in GS implemented in our model List of experiments performed for this study 20°W 0° 20°E 40°E Annual mean
4 Orbital only (MHREF)
30°N 3
0 2
20°N 1 Only includes solar insolation changes from the mid-Holocene orbit 0 Celsius 10°N 1 Slight enhancement to summer monsoon driven precipitation -1
2 -0.5 On the annual mean, mid-Holocene northern Africa colder than today 0° 3
4
10°S
Annual mean
Annual precipitation zonal mean Annual precipitation zonal mean anomaly 1000 GPCC 800 MH 30°N REF 800 PI 2000 REF MHREF 600 600
20°N 1500 ) 400 year / 400
200 mm/year mm 10°N mm\year 1000 ( 100 P 100 200 500 0° 300 0 500 0 10°S 35 30 25 20 15 10 5 0 35 30 25 20 15 10 5 0 20°W 0° 20°E 40°E Latitude Latitude 20°W 0° 20°E 40°E Annual mean
4 Orbital + Veg (MHV) 0.5 1 30°N 3
2
2 Solar insolation changes from the mid-Holocene orbit and prescribed 0 20°N 1 mid-Holocene vegetation -0.5 0 Celsius 10°N 1 -1 Further increase to summer monsoon driven precipitation
2
0° On the annual mean, mid-Holocene northern Africa becomes warmer 3 than today 4 10°S Beginning to reconcile with both precipitation and temp proxies Annual mean
Annual precipitation zonal mean Annual precipitation zonal mean anomaly 1000 GPCC 800 MH 30°N REF 800 PI MH 2000 REF V MHREF
600 MHV 600
20°N 1500 ) 400 year / 400
200 mm/year mm mm\year 1000
10°N ( 100 P 100 200 500 0° 300 0 500 0 10°S 35 30 25 20 15 10 5 0 35 30 25 20 15 10 5 0 20°W 0° 20°E 40°E Latitude Latitude 20°W 0° 20°E 40°E Annual mean
4 0.5 Orbital + Veg + Soil (MHVS) 2 1 30°N 3
3
0 2 Solar insolation changes from the mid-Holocene orbit AND prescribed 20°N 1 mid-Holocene vegetation AND changes to soil texture and composition 0.5 0 -0.5 Celsius 10°N 1 0 Further increase to summer monsoon driven precipitation -1
2
0 0° Sahara much warmer than today 3
4 Precipitation extends deeper into Sahara 10°S
Annual mean
Annual precipitation zonal mean Annual precipitation zonal mean anomaly 1000 GPCC MH 800 MH 30°N VS REF 800 PI MH 2000 REF V MH REF MHVS
600 MHV 600
20°N 1500 ) 400 year / 400
200 mm/year mm mm\year 1000
10°N ( 100 P 100 200 500 0° 300 0 500 0 10°S 35 30 25 20 15 10 5 0 35 30 25 20 15 10 5 0 20°W 0° 20°E 40°E Latitude Latitude 20°W 0° 20°E 40°E Annual mean
0 4 0 Orbital + Veg + Lake (MHVL) 0.5
30°N 1 3 0
-1 2 2
2 1 Solar insolation changes from the mid-Holocene orbit AND prescribed 20°N 1 mid-Holocene vegetation AND mid-Holocene lakes -1 0 -1 Celsius 10°N -0.5 -1 1 0 Further increase to summer monsoon driven precipitation -1 -0.5 2
0° Heterogeneity in temperature field somewhat resembling proxies -1 0 3
-0.5 -1 0 4 Slight impact on precipitation amplitude and latitudinal extent 10°S -1
Annual mean
Annual precipitation zonal mean Annual precipitation zonal mean anomaly 1000 GPCC 800 MH 30°N REF 800 PI MH MH 2000 REF VL V MHREF MH 600 V 600 MHVL
20°N 1500 ) 400 year / 400
200 mm/year mm mm\year 1000
10°N ( 100 P 100 200 500 0° 300 0 500 0 10°S 35 30 25 20 15 10 5 0 35 30 25 20 15 10 5 0 20°W 0° 20°E 40°E Latitude Latitude 20°W 0° 20°E 40°E Annual mean
0.5 4 0 0.5 0.5 Orbital + Veg + Soil + Lake (MHVSL) 0.5 0 1 3 2 2 30°N 2 3
3 0 1 -1 2 3 3 3 3 Solar insolation changes from the mid-Holocene orbit AND prescribed 20°N 0 2 1 -0.5 0 mid-Holocene vegetation AND changes to soil texture and composition 0.5 0 Celsius 10°N AND mid-Holocene lake -1 1 -1
2 0 Small further increase in precipitation compare to MHVS 0° -1 3
0 -1 -0.5 4 -1
10°S
Annual mean
Annual precipitation zonal mean Annual precipitation zonal mean anomaly 1000 GPCC MH 800 MH 30°N VS REF 800 PI MH MH 2000 REF VL V MH MH REF VSL MHVS MH 600 V 600 MHVL
) MH 20°N 1500 VSL 400 year / 400
200 mm/year mm mm\year 1000
10°N ( 100 P 100 200 500 0° 300 0 500 0 10°S 35 30 25 20 15 10 5 0 35 30 25 20 15 10 5 0 20°W 0° 20°E 40°E Latitude Latitude 20°W 0° 20°E 40°E Annual mean
0.5 4 0 0.5 0.5 Orbital + Veg + Soil + Lake (MHVSL) 0.5 0 1 3 2 2 30°N 2 3 Proxy 25 30 3 0 1 -1 2 MHVSL 3 3 3 3 20 25 20°N 0 2 1 -0.5 0
0.5 0 15 20 Celsius 10°N -1 1 -1 10 15
2 Latitude band 0 0° -1 3 5 10
0 -1 -0.5 4 -1 05 10°S 750 500 250 0 250 500 750 1000 1250 Annual mean P (mm/year) Annual precipitation zonal mean Annual precipitation zonal mean anomaly 1000 GPCC MH 800 MH 30°N VS REF 800 PI MH MH 2000 REF VL V MH MH REF VSL MHVS MH 600 V 600 MHVL
) MH 20°N 1500 VSL 400 year / 400
200 mm/year mm mm\year 1000
10°N ( 100 P 100 200 500 0° 300 0 500 0 10°S 35 30 25 20 15 10 5 0 35 30 25 20 15 10 5 0 20°W 0° 20°E 40°E Latitude Latitude Ongoing work
•Refinements of the Green Sahara land surface. East-west vegetation asymmetry, more appropriate land surface for Arabian Peninsula •Northern hemisphere vegetation change •Impact of Green Sahara on Indian Monsoon •Dynamical downscaling over Africa and the Mediterranean — connection to early human settlements Dynamical Downscaling over Asia
CAM grid with high-resolution WRF grid POP ocean grid with high-resolution ROMS grid CAM resolution ~100 km POP resolution ~100 km WRF resolution ~30 km ROMS resolution ~20 km
Yiling Huo, Chandan & Peltier (about to be submitted to CP) SST & SAT Anomalies (MHREF)
(a) (UofT-CCSM4)
20°N
10°N
70°E 80°E 90°E 100°E 110°E (b) (WRF + CROCO)
20°N
10°N
70°E 80°E 90°E 100°E 110°E
°C
-1.5 -1 -0.5 0 0.5 1 1.5
Yiling Huo, Chandan & Peltier (about to be submitted to CP) Huo and Peltier 2018, 2020, 2021 Precipitation anomalies
with reference simulation
with Green Sahara simulation
Yiling Huo, Chandan & Peltier (about to be submitted to CP) Huo and Peltier 2018, 2020, 2021 Dynamical Downscaling over Africa
•Mini physics ensemble: > 18 members •Atmosphere only and atmosphere- ocean configuration
Fengyi Xie, Chandan & Peltier (in-prep) Dynamical Downscaling over Africa
Fengyi Xie, Chandan & Peltier (in-prep) Holocene settlement data CRANE DataCRANE Project CLaSS Project: Beyond the State of the Art Settlement data (size, period, morphology) across entire region
>40,000 sites over 600,000km2
Digital ArchaeologicalUnique Atlas of dataset, the Holy Land (DAAHL)largest in the world by any metric DAAHL The CLaSS Project
CRANE: Computational Research into the Ancient Near East Summary
• Likely cause of Green Sahara was the increased precipitation driven by the West African Monsoon that was initiated by the larger northern hemisphere summer insolation, but amplified by interaction with the land surface and atmospheric dust • We have systematically examined the impact of feedback from various land surfaces on the precipitation over mid-Holocene northern Africa with a state-of-the-art coupled climate model • With all feedbacks included, modeled precipitation agrees well with proxy inferences while the modeled precipitation is sufficient to sustain applied vegetation over most of Africa. • Green Sahara leads to strengthening of the Indian Monsoon • Green Sahara conditions lead to better data-model comparison Asia • Dynamical downscaling very very important for Asia and significantly improves data-model comparison