Site selection is critical for establishing eddy covariance and robotic tram systems. These systems measure the exchanges of carbon dioxide, water and energy across the land-atmosphere boundary (eddy covariance) and the optical properties of the land surface (robotic tram system). Both technologies have specific requirements that can increase or decrease uncertainty of data. Among these are: flat terrain, homogeneous canopy height, wind speed (>1.5m/s), regional and local representation of land cover types, avoidance of current and historical disturbance and logistics. Currently there are no established quantifiable methods or standards to optimize site selection of these systems and the subsequent impact on data integrity remains unknown. We executed a multicriteria analysis following site selection models that have been applied to environmental problems and spatial missions.
Results/Conclusions
For this particular scenario, the use of ARCGIS 9 software allowed us to overlay several raster images, using a common measurement scale, and each weighted according its importance on the decision tool. The specific aims were to find a local and regional representative landscape within the study area that best fulfills instrumentation requirements, and to reduce sources of random errors by characterizing potential flux footprints variations due to changes in wind speed and direction, topography, and other environmental factors. Extensive ground surveys were required to finalize site choice. Four sites within a polygon of creosote bush (Larrea tridentata) shrubland region were selected.
GIS analysis provides a platform to build canopy height and footprint models that will allow estimation of the propagation of random errors that can be induced by environmental heterogeneity in the flux tower footprint and turbulence within it. The use of this tool is able to assist ecologists striving to make informed decisions when selecting sites for implementation of these technologies. The analysis also provides a valuable tool for assessing whether a site is representative, and how individual sites can be compared with others that are or are not being sampled.
