Top panel — cumulative heat forcing (W/m²): gross positive forcing by cause, partitioned by origin (fossil vs. ecological) and mechanism (CO₂ released vs. cooling lost). The three CO₂ bins
partition the measured atmospheric-CO₂ forcing (IPCC AR6 +2.16 W/m²) — fossil (1.68, Global Carbon Project) + ecological land-use (+0.48: soil + deforestation biomass) — rather than
summing independent bookkeeping estimates, which double-counts soil (Houghton & Nassikas 2017 bundles vegetation
and soil).
Soil CO₂ (brown): Sanderman et al. (2017), HYDE 3.2.
Deforestation CO₂, biomass only (orange): Houghton & Nassikas (2017) less its soil term.
Cooling loss (teal/blue, land + ocean): deforestation extent → W/m² via this paper's biophysical estimates (Part II table) — the novel contribution, not independently validated; ocean term ring-fenced.
Fossil CO₂ (charcoal): Global Carbon Project. The stack is gross positive forcing; the IPCC's net +2.72 W/m² is the residual after aerosol cooling (−0.5 to −1.5), not shown.
Middle panel — measured CO₂ & temperature. CO₂ (gold, ppm): EPICA Dome C (Bereiter 2015), Law Dome (Etheridge 1996; MacFarling Meure 2006), Mauna Loa (Keeling/NOAA) — the ~7 ppm Orbis dip at ~1610 is visible.
Global temperature (dark red, °C vs. 1850–1900): Marcott et al. (2013) + HadCRUT5/GISTEMP. After 2026 each curve splits into two scenarios —
emissions only vs.
+ restoration.
Bottom panel — surface energy balance (illustrative, not measured). The paper's conceptual framework:
surface heating (red) and
surface cooling (green) begin matched, then diverge as cooling capacity is lost; their gap drives a lagged
local temperature response (purple). In the restoration future, cooling rebounds within years-to-decades and local temperature falls faster than the slow CO₂ decline — biology as the fast lever.
Circled numbers (⓪–⑫) on the event line between the middle and bottom panels mark events — click the canvas or a list entry to highlight. Ocean SML degradation is genuinely uncertain.