Cloud Masking Over Ocean
If all 400 pixels in the 10km box are considered to be ‘water’ then the ocean algorithm is followed. After the gas correction described in section A1.1, the algorithm has the arduous task of separating 'good' pixels from 'cloudy' pixels. The standard MxD35 cloud mask includes using the brightness in the visible channels to identify clouds. This procedure will mistake heavy aerosol as 'cloudy', and miss retrieving important aerosol events over ocean. On the other hand, relying on IR-tests alone permits low altitude, warm clouds to escape and be misidentified as 'clear', introducing cloud contamination in the aerosol products. Thus, our cloud mask over ocean combines spatial variability tests (e.g. Martins et al., 2002) along with tests of brightness in visible and infrared channels.
The algorithm marches through the 10 by 10 box, examining the standard deviation of ρ0.55 in every group of 3 by 3 pixels. If the group of 9 pixels has a standard greater than 0.0025, then the center pixel is labeled as 'cloudy', and is discarded (Martins, et al. 2002). The only exception to this rule is for heavy dust, which may at times be as spatially inhomogeneous as clouds. Heavy dust is identified by its absorption at 0.47 µm using the ratio (ρ0.47/ρ0.65). This quantifies the difference that our eyes witness naturally. Dust absorbs at blue wavelengths and appears brown. Clouds are spectrally moderately absorbing and appear white to our eyes. If ρ0.47/ ρ0.65 < 0.75, then the central pixel of the group of 9 is identified as 'dust' and will be included in the retrieval even if it is inhomogeneous. This is a conservative threshold that requires very heavy dust in order to avoid clouds. Less restrictive thresholds would permit more dust retrievals, but might accidentally permit cloud contamination.
The spatial variability test separates aerosol from most cloud types, but sometimes fails at the centers of large, thick clouds and also with cirrus, both of which can be spatially smooth. The centers of large, thick clouds are very bright in the visible, and so we identify these clouds when r0.47> 0.40. This is an extremely high threshold that translates into an aerosol optical thickness greater than 5.0, but only for non-absorbing aerosol. Absorbing aerosol never reaches that high value of reflectance and will pass this cloud test unscathed. Some high values of non-absorbing aerosol may be discarded along with bright clouds, but this confusion is rare. Most heavy aerosol loading, with τ > 5.0, absorbs somewhat at 0.47 µm and fails to reach the 0.40 threshold value, exhibited by very bright white clouds.
Cirrus clouds are identified with a combination of infrared and near-infrared tests. Three infrared tests provided by the standard MODIS cloud mask, MxD35, are examined. The three IR tests are the “Thin Cirrus (IR) Test” (Bit 11), the “High Cloud (6.7 μm) Test” (Bit 15), and the “IR Temperature Difference Test” (Bit 18). If any of these three tests register as “applied”, then the 2×2 box of 500 m pixels (1 km MxD35 pixel) is denoted as “cloudy”, and none of these pixels are retained for aerosol retrieval. The near-infrared cirrus test is based on the reflectance in the 1.38 µm channel and the ratio r1.38 / r1.24 (Gao, et al. 2002). It is applied in the algorithm as a three step process. If ρ1.38 / ρ1.24 > 0.3, then the pixel is 'cloudy'. If 0.005 ≤ r1.38 / ρ1.24 ≤ 0.30 and ρ1.38 > 0.03 and ρ0.65 > 1.5ρray.0.65, then the pixels is also 'cloudy'. However, if 0.005 ≤ ρ1.38 / ρ1.24 ≤ 0.30 and 0.01 ≤ ρ1.38 ≤ 0.03 and ρ0.65 > 1.5ρray.0.65, then the situation is ambiguous. The algorithm labels the pixel as 'not cloudy' and will include the pixel in the retrieval process, but the quality of the retrieval (Quality Assurance Confidence - QAC) is reduced to 0, 'poor quality'. This permits aerosol retrieval at the orbital level (Level 2), but prohibits the retrieval from contributing to the long-term global aerosol statistics (Level 3). Only retrievals with QAC > 0, contribute to the Level 3 Quality Weighted products. The products and product levels are explained further in Section 3. If the reflectance at 0.65 µm (ρ0.65) does not exceed 1.5 times the Rayleigh reflectance in that channel (ρray.0.65) or the reflectance at 1.38 µm does not exceed 0.01, then the pixel is assumed to be 'not cloudy' with no ambiguity, unless the ratio ρ1.38 / ρ1.24) exceeds 0.3.
In the transition from C5 to C6, the MOD35 team identified that the IR Temperature Difference Test was falsely identifying clear ocean as cloudy in the tropics. The test was relaxed to increase clear ocean retrievals, but in testing, the aerosol team found that the MOD35 cloud mask change also increased cirrus contamination. Therefore, we lowered the ρ1.38 / ρ1.24 reflectance ratio threshold from 0.10 to 0.005 to initiate the testing for cirrus contamination. An example granule showing reduced cloud contamination in C6 as compared to C5 over ocean is shown in Figure A2.1.
Figure A2.1, Granule retrieved over the Pacific Ocean from MODIS- Aqua taken on 1 January 2010 at 22:25UTC. Top left: true color (RGB) showing scene taken from modis-atmos.gsfc.nasa.gov. (b and c) Retrieved high quality (QAC≥1 over ocean) AOD at 0.55 μm, without/with the revised 1.38 μm cloud mask test. (d) Cirrus contaminated pixels that have been removed over ocean. From Levy et al., 2013.