Stephen E. Schwartz — Brookhaven National Laboratory
Dong Huang — NASA - Goddard Space Flight Center
—
Category
Warm low clouds, including aerosol interactions
Description
Figure 1. a, RGB (red, green, blue; natural color) zenith radiance image at ARM SGP site at 16:35 UTC (10:05, local time) July 31, 2015; vertical axes denote pixel number; N, E, S, W denote cardinal directions; scale bar is evaluated for cloud height 1.5 km. b, False color image of cloud optical depth obtained from red-channel image. c, False color image of sub-region of image denoted by yellow box in a and b; note 12-fold zoom in false color scale. d, Mean values of COD calculated over 21 line profiles centered at indicated N-S pixel number bounded by pairs of horizontal lines in corresponding colors in c. e, False color image of sub-region denoted by black square in c. f, Single-pixel line profiles of COD at N-S pixel number indicated by colors of lines in e; also shown is uncertainty bar corresponding to ± 3 standard deviations of noise in retrieval of COD.
Thin clouds, which are difficult or impossible to detect and characterize by cloud radar or passive microwave radiometry, nonetheless exert quite strong radiative effects. The shortwave cloud radiative effect in the linear range, up to cloud optical depth (COD) about 3, is about 80 W m-2 per COD, somewhat dependent on solar zenith angle and surface albedo, etc. Here we report characterization of clouds at the ARM SGP site with a high-resolution commercial digital camera that obtains zenith radiance as a two-dimensional image consisting of 3456 × 3456 pixels (12 million pixels). Spatial distance is determined from angular distance in the image using cloud height determined from co-located Doppler lidar; for cloud height 1.5 km, the imaged domain is 30 m × 30 m and the nominal angular resolution of the camera, 6 µrad, corresponds to spatial resolution about 9 mm. As shown in Figure 1, downwelling zenith radiance varies substantially within a single image; radiance between successive images obtained at 4-s intervals also varies substantially, a consequence mainly of advection. Radiant intensity in red and blue channels of the camera is converted to optical depth on a pixel-by-pixel basis using a 1D radiative transfer model, with camera calibration obtained from the darkest (Rayleigh sky) and brightest (thin cloud, optical depth about 3) pixels within a series of images; because of decrease in zenith radiance with increasing COD > 3, retrieval of COD is limited at present to COD < 2. Single-pixel noise-equivalent COD is about 0.005 for COD 0.1, increasing with increasing COD. Observed spatial variation in zenith radiance is attributed to variation in COD, which exhibits considerable variation, for example, an order of magnitude within 15 m, a factor of 2 within 4 m, and 25% (0.12 to 0.15) over 14 cm. This approach, which examines cloud structure on spatial scales 3 to 5 orders of magnitude finer than satellite products and with much greater sensitivity to thin clouds than alternative approaches, opens new avenues for examination of cloud structure and evolution.
