https://press-files.anu.edu.au/downloads/press/n12084/pdf/ch09.pdf

Abstract
The earliest Homo erectus remains in
Southeast Asia are in opposite reaches
of this geographical region. H. erectus
material from Gongwangling, China, and
Mojokerto, Java, represent some of the
earliest body fossils recovered for this
species, but very few H. erectus records
exist from between these regions. We
examine possible routes that H. erectus
could have taken on their journey
southward, using a least-cost pathway
analysis. Our models suggest that the
easiest pathway ran through the centre
of Sundaland, an area now almost entirely
submerged. During periods of higher
sea-levels, however, the pathway moved
west and could coincide with emergent
areas on or just off the east coast of
Sumatra. Geological conditions on the
east coast of Sumatra, while of the right
age to contain early hominin remains, are
not conducive to the preservation of fossil
material or the retention of suitable
quarries for stone artefact production.
The Riau archipelago also lies on probable
migration routes; however, geological
outcrops of the right age will probably be
difficult to find there.