Abstract
Bi-hemispherical reflectance (BHR), in the land surface research community also known as “white-sky albedo”, is independent of the directions of incidence and viewing. For vegetation canopies, it is also nearly independent of the leaf angle distribution, and therefore it can be considered an optical quantity that is only dependent on material properties. For the combination leaf canopy and soil background, the most influential material properties are the canopy LAI (leaf area index), optical properties of the leaves, and soil brightness. When the leaf and soil optical properties are known or assumed, one may estimate the canopy LAI from its white-sky spectral albedo. This is also because a simple two-stream radiative transfer (RT) model is available for the BHR of the leaf canopy and soil combination. In this contribution, crown clumping and lateral linear mixing effects are incorporated in this model. A new procedure to estimate soil brightness is introduced here, even under a moderate layer of green vegetation. The procedure uses the red and NIR spectral bands. A MODIS white-sky albedo product at a spatial resolution of 0.05° is used as a sample input to derive global maps of LAI, soil brightness, and fAPAR at the local moments of minimum and maximum NDVI over a 20-year period. These maps show a high degree of spatial coherence and demonstrate the possible utility of products that can be generated with little effort by using a direct LUT technique.
Highlights
The correct quantitative interpretation of remotely sensed land surface reflectance data is hampered by the anisotropy of most surfaces on Earth, whether it be bare soils, vegetation canopies, or water bodies
In this paper, we investigate whether the spectral information that the NDVI is based on, namely the reflectance in the red and NIR bands, can be used to provide quantitative estimates of the leaf area index (LAI), soil brightness, and fAPAR by a simple model inversion applied to bi-spectral red and NIR white-sky albedo data, and if this is possible with a speed that is comparable to that of the NDVI calculation
In the 1D turbid medium radiative transfer model SAIL [3], an adding method [4] is applied to calculate the bi-hemispherical reflectance of the combination soil and canopy, which is given by the simple expression below [5]: rdd = ρdd +
Summary
The correct quantitative interpretation of remotely sensed land surface reflectance data is hampered by the anisotropy of most surfaces on Earth, whether it be bare soils, vegetation canopies, or water bodies This is why, after the correction of atmospheric effects, attempts are often made to remove or normalize these so-called BRF (bi-directional reflectance factor) effects. The hemispherical reflectance for incident solar radiation, called the DHRF (directional hemispherical reflectance factor) [2], is most important in this respect, since optical satellite observations are primarily only possible under nearly cloud-free conditions, so that the direct solar radiation input dominates This quantity is sometimes called “black-sky albedo” and is obviously a function of the solar zenith angle
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
