Site Selection Logic

As the cost for a full BigFoot characterization of a given site is not trivial, careful selection of sites is an imperative. The BigFoot concept requires continuous eddy flux measurements of H₂O and CO₂ exchange between terrestrial ecosystems and the atmosphere. Although the number of flux towers is growing, there are still only around 100 globally (Running et al. 1999). Biomes that cover a larger fraction of the land surface, or that are expected to experience the greatest change because of warming or land use patterns, were considered high priority sites. In addition, we selected biomes that cover a broad range in LAI and NPP to provide a robust data base for the validation of MODLand products and the examination of climatic influences on NPP.

Our boreal forest site (NOBS) represents the biosphere's second largest biome. Boreal forests have low NPP, and are interesting from the point of view that the large amount of soil carbon they contain is believed to be susceptible to release if warming occurs (Goulden et al. 1998). Our agricultural cropland site (AGRO) has high NPP, and both it and our tallgrass prairie site (KONZ) are subject to intensive land use change and management practices (e.g., cropping, burning and grazing). The BigFoot temperate mixed forest site (HARV) represents a large, historically highly disturbed biome that is purported to be a current carbon sink (Goulden et al. 1998). The desert grassland (SEVI) serves as a low anchor for LAI and NPP. The arctic tundra site (TUND) also has relatively low LAI and NPP, but it is an expected carbon source with warming. The tropical broadleaf evergreen forest in the Amazon (TAPA) has high LAI and NPP and represents the largest terrestrial biome. This biome is experiencing rapid deforestation and changes in land use. The second temperate mixed forest site (CHEQ) serves as a replicate to determine if there is variability among widely dispersed sites within the same general biome. The temperate needleleaf site (METL) is representative of xeric continental conifer forest. The five newest sites all have substantial ongoing measurement and modeling activities that complement our original efforts.

At AGRO, we are extending our field sampling for an additional year (see field activity schedule) so that we can assess MODLand product temporal valididty at this site (recall there were no MODIS data in 1999). BigFoot originally made measurements in both 2000 and 2001 at KONZ and HARV, and is continuing sampling at HARV in 2002 and in 2003. The continued work (with a less intensive field effort) at NOBS is funded by NSF. We continue working at HARV and NOBS throughout BigFoot because these two sites have long-term flux observations and are in biomes that are important to monitor under conditions of land use and global change (i.e., HARV is a purported carbon sink due to a recovery phase following widespread disturbances that took place over 100 years ago and NOBS has large carbon stores and is sensitive to climate change). Field measurements accompanying the flux data at these sites provide continued mutual corroborative support and help us understand the ecological changes that accompany the interactions of land use, climate, and carbon cycle changes. No additional field measurements are made at AGRO or KONZ beyond 2001, but using the developed remote sensing and modeling algorithms from the earlier years, we continue to monitor these sites. Sampling at the CHEQ site is 100% consistent with the BigFoot design, but at METL there will be less dense sampling than at other sites. However, the data collected and the collection protocols at METL are consistent with BigFoot.