Gravity reduction from the Earth's surface to the geoid requires knowledge
of topographical mass density. However, in practice the constant density (2
.67 g/cm(3)) is mostly used to approximate the actual density because of th
e difficulty and complexity of obtaining the actual density. This approxima
tion introduces errors in the reduced gravity, and consequently, in the geo
id. Recently, the geographical information system (GIS) was introduced as a
n efficient tool to gee-reference actual bedrock densities to digital geolo
gical maps. As a part of the effort towards the construction of the 'one-ce
ntimetre geoid' for Canada, the effects of the lateral variation of topogra
phical density on gravity and geoid were investigated in the Canadian Rocky
Mountains, Density values were estimated from the geological maps of Canad
a and the US and bedrock density tables compiled for use in the ArcView GIS
, The 5 ' x 5 ' mean and point topographical effects were computed from hei
ght and density data available on a 30 " x 60 " grid. The mean direct (topo
graphical) density effect (DDE) on gravity ranges between -4.5 and 2.3 mGal
(mean of 0.008 mGal) at the Earth's surface, and from -12.7 to 9.8 mGal (m
ean of 0.007 mGal) at the geoid. The secondary indirect (topographical) den
sity effect (SIDE) on gravity varies between -8 and 5 mu Gal. The primary i
ndirect (topographical) density effect (PIDE) on the geoid changes from -2.
5 to 1.7 cm (mean of 0.2 cm). The total topographical density effect on the
geoid ranges between -7.0 and 2.8 cm (mean of -0.8 cm). Our results sugges
t that the effect of topographical density lateral variations is significan
t enough and ought to be taken into account for the determination of the on
e-centimetre geoid.