Abstract. The Cenozoic inception and development of the Asian
monsoons remain unclear and have generated much debate, as several
hypotheses regarding circulation patterns at work in Asia during the Eocene
have been proposed in the few last decades. These include (a) the existence of
modern-like monsoons since the early Eocene; (b) that of a weak South Asian
monsoon (SAM) and little to no East Asian monsoon (EAM); or (c) a prevalence
of the Intertropical Convergence Zone (ITCZ) migrations, also referred to
as Indonesian–Australian monsoon (I-AM). As SAM and EAM are supposed to have
been triggered or enhanced primarily by Asian palaeogeographic changes, their
possible inception in the very dynamic Eocene palaeogeographic context
remains an open question, both in the modelling and field-based communities.
We investigate here Eocene Asian climate conditions using the IPSL-CM5A2 (Sepulchre et al., 2019)
earth system model and revised palaeogeographies. Our Eocene climate
simulation yields atmospheric circulation patterns in Asia substantially
different from modern conditions. A large high-pressure area is simulated over the
Tethys ocean, which generates intense low tropospheric winds blowing
southward along the western flank of the proto-Himalayan–Tibetan plateau
(HTP) system. This low-level wind system blocks, to latitudes lower than
10∘ N, the migration of humid and warm air masses coming from the
Indian Ocean. This strongly contrasts with the modern SAM, during which
equatorial air masses reach a latitude of 20–25∘ N over India and
southeastern China. Another specific feature of our Eocene simulation is the
widespread subsidence taking place over northern India in the midtroposphere (around 5000 m), preventing deep convective updraught that would
transport water vapour up to the condensation level. Both processes lead to
the onset of a broad arid region located over northern India and over the
HTP. More humid regions of high seasonality in precipitation encircle this
arid area, due to the prevalence of the Intertropical Convergence Zone
(ITCZ) migrations (or Indonesian–Australian monsoon, I-AM) rather than
monsoons. Although the existence of this central arid region may partly
result from the specifics of our simulation (model dependence and palaeogeographic uncertainties) and has yet to be confirmed by proxy records,
most of the observational evidence for Eocene monsoons are located in the
highly seasonal transition zone between the arid area and the more humid
surroundings. We thus suggest that a zonal arid climate prevailed over Asia
before the initiation of monsoons that most likely occurred following Eocene
palaeogeographic changes. Our results also show that precipitation
seasonality should be used with caution to infer the presence of a monsoonal
circulation and that the collection of new data in this arid area is of
paramount importance to allow the debate to move forward.