Monday, 11 January 2016
Greenhouse gases observing satellite (GOSAT) carries Cloud and Aerosol Imager (CAI). The CAI is a complimentary sensor for GOSAT to observe both cloud and aerosol at four wavelengths from near UV (0.38 Ám) to short infrared (1.6 Ám). The pixel resolution is 500 m except 1.6Ám channel with 1500 m. The CAI designed to observe cloud/aerosol at the same local time. We are developing the algorithm to retrieve aerosol over land using the CAI. In retrieving aerosol properties over land, surface albedo is essential. When surface albedo is contaminated by shadow, we tend to retrieve larger AOT than real world. Conventionally surface albedo was determined as minimum or second minimum of reflectance during a specific period (week/month) over each grid point in our group. However, the determined surface albedo in this way might include shadow depending on the period to determine this surface albedo over each grid. Here, we studied the direct way to identify shadow using the CAI 0.38Ám and 0.87Ám channel data. The measurements at 0.38 Ám have advantages of detection of carbonaceous aerosols and less ground surface reflection in comparison with visible bands. To take advantage of this characteristics of this channel, we compute the ratio of apparent reflectance of 0.87Ám (R0.87) over 0.38Ám (R0.38) as an indicator of shadow detection. The R0.87 becomes very small over shadow comparing with over sunny side land surface, while the R0.37 is less affected by shadow/sunny side land surface. Thus the ratio (R0.87/R0.38) becomes smaller over the shadow comparing with other scene. We found the ratio (R0.87/R0.38) smaller than 1.2 corresponds well to the shadow comparing with RGB composite image. By inspection of RGB composite image, shadow by thin cirrus can also be identified. The advantage of this method is effective for instantaneous image and regardless viewing angle. Further we can get more reliable surface albedo by discarding shadow contaminated surface within the shorter period than one month.
- Indicates paper has been withdrawn from meeting
- Indicates an Award Winner