Zhongping Lee, Xiaodong Zhang

Remote sensing reflectance (Rrs, in sr-1) is defined as the ratio of water-leaving radiance (Lw) to downwelling irradiance just above the water surface (Es). To measure Rrs, the above-water approach (AWA) is a widely adopted scheme in the past decades. In addition to the measurement of Es, AWA also involves measurements of upwelling radiance above the surface (Lt) and downwelling sky radiance (Lsky), with the latter component utilized for the removal of surface-reflected skylight (Lsky) or sky glint (Lsrs), as Lt always contains both Lw and Lsrs. In this classical paper (Applied Optics, 1999, Vol. 38, pp. 7442-7455), Mobley applied the widely-used Hydrolight to simulate Lt, Lw, and Lsky for various Sun-sensor geometries, wind speeds, and cloud cover. With Monte Carlo simulations, Mobley also showed that a roughened sea surface would reflect incident Lsky into a sensor’s field-of-view (FOV) from an area much wider than that constrained by the FOV, which greatly helped understanding the difficulties in obtaining accurate Rrs in field measurements via AWA. Based on the simulations, Mobley concluded that “until an improved method of estimating Rrs becomes available and accepted, the following suggestions can be made for using the traditional method based on Eq. (6).” This “Eq. (6)” is Eq. 4 below, and the suggestions include 1) an overall value of 0.028 for the surface reflectance (ρ) for a wind speed less than 5 m/s, and 2) an optimal viewing geometry of 40o/135o, with 40o for senor’s nadir viewing angle and 135o for sensor’s azimuth angle from the Sun. This paper has greatly impacted ocean color sciences, as these recommendations have been followed globally in the past two decades and today. However, there are a few ‘hidden stories’ in these simulations that were not disclosed or well discussed, and these hidden features put the suggestions in question.