Slow rates of rock surface erosion and sediment production across the Namib Desert and escarpment, southern Africa

Slow erosion has characterized the Namib Desert, the Namibian escarpment, a nd the adjacent Namibian highlands over the Pleistocene. Paired analyses (n = 66) of in-situ-produced Be-10 and Al-26 in quartz-bearing samples of bed rock primarily from inselbergs, of sediment from dry river and stream chann els, and of clasts from desert surfaces reveal large inventories of these c osmogenic nuclides indicating significant landscape stability over at least the past million years. Bedrock samples (n = 47) collected in three transects from the coast, acros s the escarpment, and into the highlands, show no spatial pattern in elevat ion-normalized nuclide abundance despite a difference in mean annual precip itation (MAP) between sample sites at the coast (MAP < 25 mm yr(-1)) and th ose in the highlands (MAP > 400 mm yr(-1)). Average model erosion rates inl and of the escarpment (3.2 +/- 1.5, n = 9) are indistinguishable from avera ge rates seaward of the escarpment (3.6 +/- 1.9, n = 38) indicating that ro ck on the pedimented coastal plain is eroding at the same rate as rock in t he highlands. Sediment samples (n = 3) from small streams suggest that the landscape as a whole is eroding more rapidly than the bedrock outcrops and I that a basin in the steep escarpment zone is eroding several times faster (16 m my(-1)) than either a basin in the highlands (5 m my(-1)) or a basin in the coastal plain (8 m my-1). Data from large rivers (n = 4) constrain erosion rates, averaged over 10(5) yrs and 10(4) to 10(5) km(2), between 3 and 9 m my(-1). Small quartz clasts (n = 12) collected from four desert sur faces record extraordinarily long, variable, and in some cases complex expo sure histories. Simple Be-10 model ages are as high as 1.8 my; some minimum total histories, considering both Be-10 and Al-26 and including both buria l and exposure, exceed 2.7 my. As a group, the Namibian cosmogenic data do not support the model of significant and on-going escarpment retreat. The similarity of erosion rates calculated from Be-10 analysis of fluvial s ediments and longer-term (10(7) yr), average mass removal rates estimated b y others using fission track analysis of rock suggests that Namibian erosio n rates have reached a steady state and are changing little over time. At o utcrop scales, the concordance of Be-10 and Al-26 in most bedrock samples s uggests that the model of steady, uniform bedrock erosion is valid; there i s no indication of intermittent burial, shedding of thick rock slabs, or st ripping of previous cover. At an intermediate scale, a transect of bedrock samples north of Gobabeb demonstrates that the northern boundary of the mas sive Namib Sand Sea has been steady and unshifting. Similarly low cosmogeni cally estimated erosion rates across west and central Namibia suggest that the landscape is in geomorphic steady state, its overall appearance changin g only slowly through time.