Dynamic Vegetation Simulations of the Mid‐Holocene Green Sahara

Abstract

The “Green Sahara” is a period when North Africa was characterized by vegetation cover and wetlands. To qualitatively identify the orbital‐climatic causation of the Green Sahara regime, we performed dynamic vegetation model (LPJ‐GUESS) simulations, driven by climate forcings from coupled general circulation model (EC‐Earth) simulations for the mid‐Holocene, in which the vegetation is prescribed to be either modern desert or artificially vegetated with a reduced dust load. LPJ‐GUESS simulates a vegetated Sahara covered by both herbaceous and woody vegetation types consistent with proxy reconstructions only in the latter scenario. Sensitivity experiments identify interactions required to capture the northward extension of vegetation. Increased precipitation is the main driver of the vegetation extent changes, and the temperature anomalies determine the plant functional types mainly through altered fire disturbance. Furthermore, the simulated vegetation composition also depends on the correct representation of soil texture in a humid environment like Green Sahara.

Vegetation Pattern and Terrestrial Carbon Variation in Past Warm and Cold Climates

“…Figures 4e, 4g, and 4h). In addition, simulated expansion of rainforest in equatorial Africa (Figure 1b) results from the lower temperature at the LGM (Figure S1a), which promotes the presence of evergreen forest through improved soil moisture (lower evapotranspiration; Andela et al, 2013) and a reduced prevalence of fire (Figure 2(a3), blue curve), the latter tending to favor less fire-resistant forest (Lu et al, 2018). A higher proportion of raingreen forest and grassland in experiment "LGM−TEMP" (Figure 4b) reflects these dominant mechanisms.…”

Impacts of Large‐Scale Sahara Solar Farms on Global Climate and Vegetation Cover

“…In the desert, vegetation growth enabled by rainfall enhancement further reduces albedo, increases evapotranspiration, and decreases sensible heat flux; this reinforces the initial precipitation increase and leads to a larger vegetation response (Li et al., 2018). As a result, the simulated vegetation expansion, which in the Sahara mostly consists of grass replacing bare ground, induces positive land (vegetation)–atmosphere feedbacks (Lu et al., 2018; Pausata et al., 2016). This local positive albedo–precipitation–vegetation feedback is also known as the classic Charney mechanism (Charney, 1975).…”

Reconstructing Past Global Vegetation With Random Forest Machine Learning, Sacrificing the Dynamic Response for Robust Results

“…These dynamic vegetation simulations were run at a resolution of ∼1°, at a 1-h time step. The baseline simulations for the PI, MH, and LGM in terms of the vegetation distribution are well discussed in previous studies (Lu et al, 2018(Lu et al, , 2019. The offsets in the perturbation simulation results (+spread & −spread in Figures 3, 4, and 5) from the baseline simulations are compared with the RF approach to test its response.…”

“…One method to produce such vegetation maps, regardless of information from reconstructions, is to perform offline dynamic vegetation model simulations (e.g., Lu et al, 2018Lu et al, , 2019; I. C. Prentice et al, 2011). However, available models to reconstruct or predict vegetation from climate rely on absolute values of temperature, precipitation, and a range of other climate variables.…”