Vegetation water content (g of water per m2) is an important parameter for modelling land surface emission and backscatter at L-band. Therefore, destructive vegetation samples were collected at various locations in the study area, with the aim to characterize all the vegetation types present. Surface spectral reflectance and leaf area data are valuable in developing methods to estimate the vegetation water content and other canopy variables over large areas. Spectral observations were made concurrent with biomass sampling in order to provide the essential information needed for larger scale mapping using aircraft and satellite observations. During SMAPEx-4, a Multi Spectral Radiometer (MSR) developed by CROPSCAN was used. Note that the MSR bands used coincide with the MODIS instrument. Leaf area was measured using a LAI-2000.
The vegetation sampling strategy was based on the assumption that the changes in
vegetation (biomass, VWC and plant structure) are negligible within a week, and therefore
the same paddock was sampled with one week revisit time.
Vegetation samples for biomass, vegetation water content, soil surface reflectance and LAI
measurements were collected daily at the 2.8km × 3.1km focus areas. Vegetation sampling largely followed the sampling schedule of the soil moisture monitoring. However, since cropping areas (YA4, YA7 and YF) are expected to present
a large variety of vegetation types and growth stages to be sampled, as opposed to the more
uniform dry land areas (YB5, YB7 and YE), the former will be strictly sampled when coincident
aircraft spectral observation of the area are scheduled.
Vegetation sampling was repeated at the SAME locations as the week before, to
accurately track temporal changes in vegetation biomass. Within each focus area, all major vegetation types were monitored. In the eventuality that
different growth stages of the same vegetation type existed within the sampling area, they were independently sampled. Each major vegetation type (or growth stage of the same vegetation type) was
characterised by making measurements at a minimum of 5 sampling locations distributed
within “homogeneous” crops/paddocks. Additional vegetation sampling was performed outside the focus areas when a major
vegetation type observed within the SMAPEx-4 study area was not represented in these. All vegetation measurements were prioritised between approximately 10am and 2pm
eastern standard time to optimise the ground spectral observations. To assist with interpolation of vegetation water content information and derivation of a land
cover map of the region, the vegetation type and vegetation canopy height were recorded
for each vegetation type sampled. In the case of crops, additional information on row
spacing, plant spacing and row direction (azimuth angle) were recorded.
Vegetation sampling was concentrated in the 3km x 3km focus areas. Within each area, the major vegetation types and phonological stages were characterised by making measurements at 5 locations distributed within homogeneous crops/paddocks. At each location, measurements consisted of 1 destructive sample, 54 LAI and 25 CROPSCAN readings. LAI and CROPSCAN were then averaged, after quality control, into a single LAI and CROPSCAN reading per location. Additional observations at each location included plant height as well as crop row spacing and direction and plant spacing in crops. Complete details on the vegetation sampling strategy and protocols can be found in the SMAPEx-4 workplan.