mksrf_soilcol.081008.nc                           Erik Kluzek

Dataset from Peter Lawrence created on Oct/10/2008

The new CLM raw dataset has been developed to address high grass 
contributions in the current CLM 3.5 raw datasets and high surface 
albedo in grasslands particularly in the near infrared. The high grass 
contribution was found to be coming from the use of the MODIS Vegetation 
Continuous Fields data to split vegetation fraction to tree and 
herbaceous components. This appears to work well in grasslands, 
shrublands and savannas, however performs poorly in forested areas. This 
is especially true in broadleaf deciduous forests where the seasonality 
of the greeness has been taken to show large amounts of herbaceous 
foliage in the VCF data. To overcome this problem we over write the VCF 
breakdown between trees and herbaceous foliage in forests so that the 
PFTs of a forest now are trees (tree + herbaceous VCF) and bare ground 
(bare VCF).

The high albedo over grasslands was addressed with new field measured 
grass optical properties from:

Asner, G.P., C.A. Wessman, D.S. Schimel, and S. Archer. 1998.
Variability in leaf and litter optical properties: implications for 
canopy BRDF model inversions using AVHRR, MODIS, and MISR.
Remote Sensing of Environment 63:200-215

In their paper they found the optical properties were much less 
reflective in both the visible and near infrared than the Dorman and 
Sellers values which have been used in a wide range of land surface models.

The changes in grass contribution and optical properties greatly reduce 
the remaining albedo biases in CLM. The changes in PFTs also improves 
the albedo signal in land cover change experiments where forest are 
replace with cropping or grasslands.