Comparison of the spectral albedos from Triana against model calculations will require an extensive modeling infrastructure. The model calculations will require information on the cloud physical and microphysical properties, atmospheric thermodynamic profiles, profiles of ozone, and spectral surface albedos. In the analysis of Nimbus-7, the cloud and atmospheric properties were obtained from the NCAR GCM and from the ISCCP data sets. For Triana, equivalent data sets can be obtained by combining data from Earth Observing System (EOS) satellites, from operational meteorological centers, and from the Triana cloud products. In order to calculate spectral albedos as measured by Triana, it is necessary to specify spectral albedos in the same wavelengths at the Earth's surface. The Nimbus-7 study was limited to ocean regions since spectral albedos for the ocean surface have very small uncertainties. The visible and near-infrared albedos for land surfaces and the polar caps are much less certain. Observational data is limited to small geographic regions, and the spectral albedos calculated with land-surface models (for example, the land models used in coupled GCMs) have not been validated over most of the globe. Since the Nimbus-7 spectral radiometers had wide field-of-view apertures, spectral land surface albedos cannot be inferred using data from these instruments. However, for Triana it will be necessary to model radiative transfer over land and ice-covered surfaces as well as over ocean regions. Work is already underway for characterizing surface albedo at near-infrared wavelengths, 870 and 1600 nm, from AVHRR and VIRS (e.g. Sun-Mack et al., 1999). The effort to measure surface albedos over enough near-IR wavelengths to derive relatively accurate values globally will continue using additional channels on MODIS and other Terra instruments.
A strategy will be developed for analyzing the Triana data using the modeling infrastructure developed for the Nimbus-7 measurements. This strategy can be thoroughly tested before the launch of Triana. Like Triana, Nimbus-7 also imaged the entire disk of the planet as viewed from the orbital altitude. Thus the analysis techniques originally developed for Nimbus should be applicable to Triana. The first step is to compute spectral albedos for land and ice from a global land-surface model (LSM). We will use the NCAR LSM (Bonan, 1998) or comparable model to generate these albedos. The albedos will be generated from a simulation of the Nimbus time period using observed sea-surface temperatures. The LSM, like many models of its class, includes a realistic time-varying prescription of the types and geographic distribution of vegetation (Bonan, 1996). The comparison of Nimbus-7 data against model calculations will be extended to land and ice surface using albedos from the LSM. We will then test the sensitivity of the results by varying the spectral albedos within a reasonable range of uncertainty. The spectral and broadband albedos from the land-surface model will be also evaluated with data from CERES (broadband only), field observations, and the MODIS data. Extensive MODIS clear-sky spectra should become available for oceanic and continental regions during late 2000. It should also be possible to compare just the Triana data for the Pacific Ocean basin against the model. This comparison should be minimally affected by uncertainties in the land-surface albedos.
Once such a modeling infrastructure is available, the possibilities for further tests of radiative transfer models become great, not just with NISTAR but also with EPIC. Because of Triana¹s simple viewing geometry and relatively simple data processing requirements compared to LEO satellites, scientists and students would be able to study a wide variety of phenomena without many of the complexities usually associated with remote sensing. Because of the lunar calibration for EPIC and absolute calibration for NISTAR, the scientific community would be able to focus on geophysical applications of a stable, high-accurate data set. This could have important repercussions both for remote sensing and climate.