Light radiation striking a vegetative canopy interacts with the elements of the canopy (leaves, stems, branches) and the underlying substrate. The nature of the interaction depends on the wavelength of the light, the light quality (direct or diffuse), the illumination incident angle, the canopy component optical properties, and the canopy architecture. Radiation is reflected, transmitted, or absorbed by the canopy, and at in the thermal wavelengths radiation is also emitted. How radiation interacts with the canopy provides information on the characteristics of the canopy. The surface radiation group collected information on how radiation was reflected, emitted, and transmitted through plant canopies to gain an understanding of the biophysical characteristics of the canopies.

The surface radiance data set consists of broad-band spectral bidirectional reflectance data (PARABOLA, MMR_GRND_DATA, MMR_HELO), and high spectral resolution bidirectional reflectance data (SE5_UNL, SE5_GSFC, SE5_HELO, GEM_HELO) from both helicopter-based and ground observations. The optical properties of canopy components such as leaf reflectance and transmittance in both broad spectral bands (MMR_LEAF) and high spectral resolution (SE5_LEAF), and soil reflectance (SOILREFL) are also included. Canopy structure can be inferred from light transmittance data (LB_KSU, LB_UNL, LIGHTWND).

In the thermal bands emitted radiation from canopy elements becomes important. Thermal infrared measurements were made on the ground (SURF_RAD, IRT_MULT, MMR_GRND, IRT_GRND) and from the helicopter (IRT_HELO, MMR_HELO). For calibration purposes of the airborne thermal measurements radiant temperatures were collected for water targets (WATEMP87, WATEMP89).

The radiation environment at the surface was measured, including the incoming and reflected short-wave radiation, the albedo, net radiation, air temperature, and humidity (SURF_RAD). The amount of incoming radiation in the MMR bands throughout selected days is provided in MMR_CALB. Incoming long-wave radiation was measured in LONG_RAD.

The data from the Portable Apparatus for Rapid Acquisition of Bidirectional Observations of Land and Atmosphere (PARABOLA) were collected to measure variations in the vegetation reflectance as a function of solar and sensor viewing geometry, wavelength, and plant canopy biophysical characteristics. The PARABOLA consists of a rotating head mounted on a boom which allows the head to freely rotate in all directions. Actual data values are bidirectional reflectance factors averaged into bins of standard view and zenith angles, in three wavelength bands (0.65-0.67, 0.81-0.84, and 1.62-1.69 micrometers) under various solar conditions throughout the day. PARABOLA measurements were made during each of the five FIFE Intensive Field Campaigns (IFCs) from six sites. These data were collected by D. Deering (NASA Goddard Space Flight Center).

The PARABOLA consists of a rotating head mounted on a boom which held the instrument 4.5 m above the ground and allowed the instrument to scan the entire sphere, thus both sky and ground radiances are observed. Observations are made in three wavelength regions (0.65-0.67, 0.81-0.84, and 1.62-1.69 um). The actual data values are bidirectional radiances and reflectance factors averaged into bins of standard view and zenith angles. The PARABOLA design provides multidirectional viewing, but the geometry of the system does not allow the same spot on the ground to be measured in each view direction. Thus, target surfaces that are homogeneous over relatively large areas are required to compare the different bidirectional view angles. The 15-degree IFOV results in sampling of spots ranging from 2 square meters at-nadir to 17 square meters at 60 degrees off-nadir. PARABOLA measurements were made during each of the five FIFE Intensive Field Campaigns (IFCs) from six sites.

The Barnes Modular Multiband Radiometer (MMR) collects data in eight wavebands (0.45-0.52, 0.52-0.60, 0.63-0.69, 0.76-0.90, 1.15-1.30, 1.55-1.75, 2.08-2.35, 10.4-12.5 micrometers). Wavebands 1 through 4 have silicon detectors, Wavebands 5 through 7 have lead sulfide detectors and Waveband 8 has a Lithium Tantalum trioxide detector. The MMR provides analog voltage responses to scene radiance in each waveband. The voltages can be converted to radiance by using calibration coefficients developed from least squares fitting of integrating sphere observations under varying light levels. Reflectance factors can be calculated by two different methods. If a single MMR is used, sequential observations from both the target and calibration panel can be directly compared. The target observation is divided by the calibration panel observation to obtain the reflectance factor. The second method involves using two instruments where one collects data from the target surface and the other collects data from the calibration panel. The data from both instruments are first converted to radiances and then ratioed.

A calibration panel is a surface whose reflectance properties are well characterized and can be used to compare with the target observations. In FIFE the calibration panels were halon or barium sulphate painted metal plates. These two substances are good diffuse reflectors; however, sun angle adjustments must be made to account for the fact that these are not perfect diffuse reflectors.

The Nebraska Multiband Leaf Radiometer (NMLR), used to measure leaf hemispherical reflectance and transmittance is based on a modified MMR and makes measurements in the same wavebands. The NMLR was attached to a Li-Cor integrating sphere with an external light source with the beam restricted to only illuminate the leaf.

The Spectron Engineering SE590 spectroradiometer has a detector head that uses a diffraction grating at the dispersive element; the spectrum is imaged onto a 256-element silicon photodiode array. Each element integrates simultaneously acquiring the spectrum in a fraction of a second. Each channel is approximately 0.003-micrometer wide with the spectral range of the instrument being from 0.372 to 1.1 micrometer. At the ends of the spectral range of the SE590 there is an increase in the noise in the data due to decreasing sensitivity in the silicon detectors and decreasing amounts of incoming light at those wavelengths. Note that each SE590 has slightly different radiometric responses and wavelength ranges for each channel. The wavelength of each SE590 channel is determined by observing illumination sources with distinct known spectral peaks, noting in which channels those peaks occur, and using a spline fit from these known points to calculate the wavelength of all channels.

The SE590 is able to collect a spectrum in about a second. The original data values are digitized counts. The counts can be converted to radiance by using calibration coefficients developed from least squares fitting of integrating sphere observations under varying light levels. Reflectance factors can be calculated by two different methods as described for the MMR. If a single SE590 is used the target and calibration panel can be directly ratioed to obtain the reflectance factor. If two instruments are used, as in the case of the helicopter, one collects data from the target surface and the other collects data from the calibration panel. The data must first be converted to radiances followed by adjustments for differences in the wavelength calibrations of the instruments. To make these wavelength adjustments usually a spline fit is made to the radiance data, then the data are resampled at even intervals. The resampled radiances can now be ratioed to calculate reflectance factors.

The Everest IRT model 4000, used to collect the IRT_GRND and IRT_HELO, measures radiation emitted from a surface in the 8- to 14-micrometer wavelength region. The emitted thermal radiation is related to the surface temperature by the Stefan-Boltzmann equation. The IRTs were calibrated against a precision black-body source.

The University of Nebraska at Lincoln group made bidirectional reflectance and surface temperature measurements using the MMR, SE590, and Everest IRT. In each case the instrument head was mounted at the end of a portable 3.4-m-tall mast. One end of the mast was placed on the ground and the mast was manually tilted through a series of view zenith angles. All of these instruments have a 15-degree field-of-view lens and in each set of measurements observed the same point on the ground, although the area viewed varied with view zenith angle. At a nadir view, a circle with a diameter of 0.75 m was observed. Most measurements were made in the plane containing a vertical from the surface and the sun, known as the solar principal plane.

As noted above, each SE590 has a unique wavelength associated with each of its channels. So that comparisons can be made among SE590s the UNL group applied a cubic spline interpolation to the SE590 spectrum to standardize the wavelengths to every 0.005 micrometers from 0.4 to 1.0 micrometers. In 1989 downwelling long-wave radiation was calculated from the detector temperature information from the UNL MMR during periods of bidirectional reflectance measurement collection.

The Goddard Space Flight Center (GSFC) SE590 was mounted on the boom with the PARABOLA instrument. The boom held the instrument 4.5 m above the ground. The SE590 was equipped with a 15-degree field-of-view lens. The data are nadir views collected at approximately every 10-degree change of solar zenith angle to characterize diurnal variations in canopy reflectance. The spectral measurements in this data set are the standard SE590 output of 0.372 to 1.1 micrometers at intervals of approximately 0.003 micrometers. Several scans were averaged together.

During the 1987 field campaigns only the MMR was flown on the NASA Bell UH-1B helicopter. In the 1989 field campaign, a SE590, an Everest IRT, and the Gemma spectroradiometer were added. The helicopter was able to visit most of the FIFE sites during a flight. At the sites, the helicopter would hover or slowly fly at approximately 330 m above ground level while the instruments collected data. These data were averaged together to determine site averages. The spatial coverage of these averages was a ring around the site center in 1987, and within the Wind Aligned Blob (WAB), an area upwind of a flux tower, in 1989. The MMR and SE590 were equipped with a 1-degree field-of-view lens which resulted in a circle with a diameter of 5.76 m observed in a nadir view. The IRT had a 4-degree field of view which resulted in a 23-m-diameter circle viewed. The Gemma had a variable focus lens system between 80 to 200 mm yielding a FOV between 3.6 x 0.3 degrees and 1.5 x 0.1 degrees.

Off-nadir backscatter data in the solar principal plane were collected for a few sites for the MMR, SE590, and IRT. The Gemma mount was fixed at nadir.

A second MMR and SE590 on the ground collected calibration panel data during the helicopter flights. Radiance values were calculated from the helicopter and calibration panel data. For the MMR, the calibration panel radiances were adjusted for the effects of variations in solar zenith angle on the reflectance of the calibration panel. These data are available in MMR_CALB. For the SE590, a cubic spline interpolation was applied to the radiance values to standardize the wavelengths to every 0.005 micrometers from 0.4 to 1.1 micrometers, and spectral reflectance was calculated from the ratio of the helicopter and calibration panel data. The SE590 calibration panel radiances were not adjusted for solar zenith angle effects. Reflectance factors were not calculated for Gemma radiance data.

The PARABOLA measures bidirectional reflectance in three broad spectral bands. Other broad-band bidirectional reflectance data collected were from the Barnes Modular Multiband Radiometer (MMR). MMR bidirectional reflectance data sets were collected by the UNL group using their mast, sometimes at sites near to the PARABOLA. These data are available in MMR_GRND. MMR data were also collected from the helicopter. These data cover more sites, but there are few off-nadir views available. The helicopter MMR data are available in MMR_HELO. Another broad-band instrument used was the Exotech. Nadir-viewing Exotech data of the mowing experiment (see Surface Biology description) were collected.

PARABOLA data were also collected at the same sites as the SE5_GSFC data.

The SE590 Spectroradiometer was used in four data sets. The UNL and the GSFC data sets are bidirectional reflectance data collected on the ground. They differ in collection methods with the UNL SE590 mounted so that it viewed the same area of the ground as view angle changed, while the GSFC SE590 rotated atop a boom resulting in quicker data collection, but viewing different areas with each view angle. The GSFC SE590 data were collected at the same location as the PARABOLA data. The UNL SE590 data were collected in conjunction with several other measurements. Incoming and reflected short-wave radiation and net radiation were measured by the UNL group at their sites. Light bar and light wand data were gathered at the UNL study plots as well as destructive samples of biomass and LAI. The destructively sampled data are in BIOLOGY\VEG_BIOP with a PI name of B. Blad.

The helicopter-mounted SE590 viewed a larger area than the ground-based instruments and it was able to visit more sites. Also mounted on the helicopter were an MMR and an IR thermometer.

SE590 leaf optical properties data were collected for the most common species at three sites in 1989. A more extensive data set of broad-band leaf spectral reflectances and transmittances were collected by the UNL group using a modified MMR.

Biophysical parameters such as leaf angle and LAI can be estimated from the light bar and light wand data. These data can be compared with direct measurements of these variables. The data set BIOLOGY\VEG_BIOP contains destructively sampled biomass and green LAI measurements, BIOLOGY\LEAF_ANG contains direct measurements of leaf azimuth and inclination angles (see the Biological Characterization summary document).

Broad-band (8-14 microns) infrared radiometers were used to collect data both from the helicopter and on the ground. Surface temperature data collected on the ground at several different view zenith angles are available in IRT_GRND and IRT_MULT. Also, ground-measured surface temperature data are in SURF_RAD. Surface temperature data were also collected from the helicopter. These data cover more sites and are found in IRT_HELO. Note that most of the MMR data also reports surface temperatures, and that incoming long-wave radiation was calculated from the MMR detector temperatures by the UNL group.

The helicopter-mounted MMR and SE590 used a second ground-based MMR and SE590 to collect calibration panel data. These data were combined to calculate reflectance factors. There was no correction for differences in the illumination between the calibration panel site and the site where the helicopter was collecting data. For the SE590 calibration panel spectra were collected approximately every 15 minutes during a helicopter flight. These data were then linearly interpolated to produce calibration panel data at the time of the helicopter data collection. The calibration panel data were not corrected for varying sun angles. No attempt was made to correct the helicopter data for the effects of the atmosphere between the ground and the helicopter.

The soil spectra from Stoner and Baumgardner were collected in a laboratory using sieved samples with soil moisture controlled. None of the soils reported were collected at the FIFE site, the spectra reported are from soils with similar physical and chemical surface properties. Note that soil in situ will have different optical properties due to differences in structure and moisture content. Also, the actual background to the prairie canopies will include dead and decaying leaves, especially for the unburned sites.

The algorithms used to calculate biophysical parameters using transmitted light make assumptions about the characteristics of the canopy, such as a random distribution of leaves and leaf azimuth angles. The UNL and Staff groups light bar calculations for LAI use different equations making different assumptions about the leaf inclination angle distribution in the canopy. The light wand uses yet another algorithm to calculate LAI and mean leaf inclination angle.

In the light bar data, measurements were not made of the amount of light reflected from the background back into the canopy. This value was assumed to be small and is not included in calculations of Fapar. Some of the Staff light bar measurements were made early and late in the day resulting in large solar zenith angles. Data collected at these times give a much higher Fapar value than midday values would.

The UNL group corrected the measured radiance for surface emissivity and reflected long-wave radiation using measured values for the latter two quantities. Therefore, their results are not directly comparable to brightness temperatures measured from an aircraft or satellite platform unless the same corrections can be applied to the latter data, which is not always possible.

The surface reflectance investigations can be grouped into four areas: (1) measurement of bidirectional reflectance and estimation of hemispherical albedo; (2) evaluation of the spatial and seasonal variability of reflectance and vegetation indices; (3) determination of surface and radiation characteristics and their effects on vegetation indices; and (4) measuring surface temperatures for estimating sensible heat flux.

Bidirectional reflectance measurements were made on the ground with the SE590, MMR, and PARABOLA, and from the air using the Advanced Solid-state Array Spectroradiometer (ASAS) and the helicopter-mounted MMR and SE590. The bidirectional reflectances from all of the instruments were compared and found to agree well qualitatively, capturing the key spectral effects. They also agreed reasonably well quantitatively, but consistent biases between instruments and groups were observed. These biases may be due to differences within sites where the data were collected and the variation in scales observed by different instruments using different observing platforms and methodologies. Reflectances from the airborne sensors were slightly higher than the ground-based instruments. Variation in solar zenith angles can also effect the bidirectional reflectance pattern. The best comparisons between instruments occurred at large solar zenith angles.

The helicopter was well suited as a platform for sampling the spatial and temporal variability of the surface reflectances. Multiple sites were observed in a flight usually lasting about two hours and flights were made during all IFCs. This compares with the ground-based groups which were able to visit only two or three sites a day at most. The helicopter MMR data show that while all eight bands contribute information useful to characterizing spatial and temporal variations in surface reflectance, MMR Bands 7, 3, 6, and 1 were most sensitive to spatial variations, and MMR Bands 3, 7, 8, and 1 were most sensitive to seasonal changes. Surface reflectance was found to be affected by green leaf area, and management practice, such as burning and grazing. The Simple Ratio vegetation index was found to be a linear function of LAI.

The surface reflectance is affected by the optical and biophysical properties of that surface. During FIFE, measurements were made of these properties to gain a better understanding of how to use remote sensing to derive biophysical variables. Leaf reflectance and transmittance data, soil reflectance, bare-ground reflectance, leaf angle distribution data, green leaf areas, standing live and dead biomass, and fraction of absorbed PAR (Fapar) data were all collected during FIFE. These data have been used to parameterize and check canopy reflectance models such as the SAIL model and the Ahmad-Deering model. For example, the SAIL model was found to give good agreement with the helicopter MMR and light bar data, and it indicated that the Normalized Difference Vegetation Index (NDVI) was related to the fraction of PAR absorbed by the green biomass.

One interest of FIFE was the use of infrared thermometers to determine the sensible heat flux. One method of estimating the sensible heat flux uses the temperature gradient between the surface and the air. This method requires an accurate measurement of the surface temperature. The observed surface temperature of a canopy varies with the view angle of the infrared thermometer. It was found that a nadir view provided poor estimates of the canopy surface temperature for the sensible heat flux calculations. Off-nadir views give better results but the exact angle to view was shown to vary with wind speed.

Ahmad, S.P., and D.W. Deering. 1992. A simple analytical function for bidirectional reflectance. J. Geophys. Res. 97(D17):18867-18886.

Blad, B.L., and D.S. Schimel. 1992. An overview of surface radiance and biology studies in FIFE. J. Geophys. Res. 97(D17):18829-18835.

Deering, D.W., E.M. Middleton, J.R. Irons, B.L. Blad, E.A. Walter-Shea, C.J. Hays, C.W. Walthall, T.F. Eck, S.P. Ahmad, and B.P. Banerjee. 1992. Prairie grassland bidirectional reflectances measured by different instruments at the FIFE site. J. Geophys. Res. 97(D17):18887-18904.

Hall, F.G., K.F. Huemmrich, S.J. Goetz, P.J. Sellers, and J.E. Nickeson. 1992. Satellite remote sensing of surface energy balance: success, failures, and unresolved issues in FIFE. J. Geophys. Res. 97(D17):19061-19090.

Vining, R.C., and B.L. Blad. 1992. Estimation of sensible heat flux from remotely sensed canopy temperatures. J. Geophys. Res. 97(D17):18951-18954.