Description:
The dataset contains the MODIS cloud data that has been analyzed in Sengupta et al (2012) and Sengupta et al (2016). The variable that was modeled are the final Q-values which reflect the confidence of clear sky for a pixel located at the specified latitude and longitude values. For more details about the MODIS cloud data please see Sengupta et al (2016).
Format:
The file mod35_2006180_1245_test_output.out contains the following columns, in this order, for each pixel:
- scan line number
- pixel number
- surface type
- latitude
- longitude
- initial Q values
- final Q values
The Surface types are coded as follows:
- 0 = water
- 1 = coast
- 2 = desert
- 3 = land
Sengupta et al (2016) also made use of a sun-glint indicator variable. This is provided in the file SG.mat.
References:
-
Sengupta, A., Cressie, N., Frey, R., and Kahn, B. H. (2012). "Statistical modeling of MODIS cloud data using the Spatial Random Effects model." In Proceedings of the 2012 Joint Statistical Meetings, Alexandria, VA: American Statistical Association, 3111-3123
-
Sengupta, A., Cressie, N., Kahn, B. H., and Frey, R. (2016). "Predictive inference for big, spatial non-Gaussian data: MODIS cloud data and its change-of-support." Australian & New Zealand Journal of Statistics. 58, 15–45.