4 - DESCRIPTION OF MEASUREMENTS 
4.1. Directional Temperature 
Measurements of directional brightness temperature were made at the First ISLSCP Field 
Experiment (FIFE) using a Barnes Modular Multiband Radiometer (MMR) mounted on a 
hand-held mast that could be tilted in various directions. The site and sampling procedures 
are described in Starks et al. (1991). An area of somewhat less than one square meter was 
viewed from nadir to 50° zenith angles and measurements typically were made on about eight 
plots at a site. Data and documentation are available on CD-ROM (Strebel et al., 1992). 
4.2. Directional Emissivity 
Measurements of directional emissivity at the FIFE experiment were made by Palluconi and 
Kahle (1990). 
Measurements of directional emissivity are very difficult to make and frequently require 
assumptions about equality of emissivity in several wavelength bands (Palluconi and Kahle, 
1990). In a effort to overcome these limitations, we have developed a new method for esti 
mating directional emissivity of soil-vegetation surfaces, which we shall refer to as the 
"atmospheric cavity method." Essentially, directional radiance or temperature measurements 
are made in two wavelength bands: The first band is in a region of atmospheric transparency 
such as 8-14 or 10-12 pm; the second band is in a region of atmospheric absorption such as 
14-22 pm. If both infrared thermometers (or radiometers) are calibrated to each other with 
sufficient precision, then the measurement in the atmospheric window provides a brightness 
temperature for a particular wavelength band, and the measurement in the region of atmo 
spheric absorption provides an estimate of directional infrared temperature, which is an 
approximation of the surface kinetic temperature. Even though the emissivity of the surface 
in the atmospheric absorption band is not unity, the directional temperature approximates the 
kinetic temperature because the atmosphere and target surface are at nearly the same tempera 
ture; thus any flux from the sky that is reflected from the surface in this wavelength band 
originates from nearly the same temperature as the surface itself. 
Four Infrared thermometers (Model 4000, Everest Interscience, Inc., Fullerton, CA, USA) 
are calibrated to each other within 0.1 C and mounted on a device that will point them at the 
surface and also at the sky at precise angles (within one degree). Because of drifts in the 
infrared thermometers, they were calibrated using a black body every minute to insure consis 
tency within 0.1 C. Two infrared thermometers (one 8-14 and one 14-22 pm) viewed the 
canopy or soil and two viewed the sky. Measurements were made at 0, 15, 30, 45, 60, and 
75° from nadir at two azimuths 180° apart, and between each measurement at an angle to the 
surface a nadir measurement was taken to insure that no significant changes in surface tem 
perature were occurring. Approximately five minutes were required to make a complete set 
of surface measurements. Sky temperature in both wavelength bands was also measured at 
the same six angles. Measurements are begun in late afternoon under clear skies and con 
ducted until after sunset. During this period, the sky temperature in the 14-22 pm band varies 
from below surface temperature to above surface temperature so that the sky can be treated 
as a large black-body cavity of nearly the same temperature as the surface. In addition, dew 
usually has not formed at this time of the day so the canopy and soil are dry. 
In the laboratory, two factors must be characterized: 1) The dependence of the 14-22 
P™ infrared thermometer temperature on C0 2 concentration and the difference between air tem 
perature and target temperature, and 2 ) infrared thermometer temperature corrections