742 
returning to a savannah of herb and grass understorey with scattered Guiera senegalensis shrubs, and (3) tiger- 
bush, a type of natural open forest which contains lenticular patches of vegetated cover containing two species 
of shrubs ( Combretum micranthum and Guiera senegalensis)) with a single tree species ( Combretum nigricans) 
within large areas of bare soil. Table 1 indicates the dates during the IOP when ATLAS made surface and/or 
atmospheric measurements and their relationship to overpasses by the Landsat 5 Thematic Mapper sensor or 
airborne sensors on the NASA C-130 including the Advanced Solid-State Airborne Spectroradiometer (ASAS) 
which records multiple for- and aft- look directions in the flight azimuth at a given area of the ground. It should 
be noted that the 17th September 1992 combined overpasses by both aircraft and satellite sensors with 
coincident ATLAS measurements at the Tiger-Bush sub-site and represents the main 'Golden Day 1 of 
observations for the surface radiation budget experiments. 
Table 1. Timetable of ATLAS deployments and remote sensing overpasses during the Intensive Observing 
Period (IOP). 
2.3 Measurements 
The standard measurement procedure employed in Niger was to set up ATLAS on the boom above the soil 
surface or vegetation canopy of interest and measure either a full set of directional surface reflectances (BRDFs) 
or a combination of measurements within the Solar Principal Plane (SPP) or at 45” to the SPP to match the flight 
azimuths used during the airborne overpasses. This was done using the 15‘ FOV head with near simultaneous 
measurement of global irradiance from the cosine head. Where possible these measurements were duplicated 
with minor displacements of the boom to average the recorded reflectances from a given direction and also 
repeated at intervals throughout the day (weather permitting) to assess the effects of various solar zenith angles. 
Full sky scans using the 1 ° FOV were interspersed to retrieve a measure of the diffuse sky anisotropy. A major 
problem was encountered with the ATLAS instrument on its first use in deriving direct solar beam irradiance. It 
was noted from the retrieved values that there was a s mall , but unknown, displacement of the optical alignment 
between the 1’ FOV spectroradiometer and the solar tracker thus obviating the precise pointing capability 
required to make measurements from the ~0.5‘ solar disk. It was not possible to correct for this error while out in 
Niger and a large proportion of the data that the instrument was designed to measure is unavailable, notably the 
atmospheric optical depths. Despite these problems some 7,500 spectra were collected during the IOP and 
include BRDFs from examples of the characteristic cover-types and a good sample of the diffuse sky irradiance 
under different atmospheric loadings and at various solar zenith angles. 
2.4 Database Construction 
A database has been designed which has some 267 columns of data, including the 252 spectral values for each 
spectrum measured as illustrated in Table 2. Each row of the database will contain a spectrum of either sensor 
received radiance from the 1" or 15" FOV heads, irradiance from the cosine corrected head, or for spectra from 
the ground hemisphere a set of calculated surface reflectances. As the database is filled with calibrated and 
verified spectra, groups of associated spectra can be addressed covering one of the pre-defmed observation 
modes of the instrument. Table 2 is representative of the first four spectra in a set of Solar Principal Plane 
measurements (Obs_Mode = 6) that have been converted to surface reflectance. The full complement of 
instrument observation modes are given in Table 3.