Earthdata Forum

I want to use MODIS L1B data to verify the cloud detection algorithm. I downloaded the MOD021KM data, and I want to extract the reflectance of 1380nm channel in EV_1KM_RefSB. I know that the value read from the table multiplied by reflectance_scales is the reflectance value I finally need. But I am not sure whether reflectance_offset has effect on reflectivity. If so, what should I do?

I also want to match the L1B data with the L2 level product data. I find that the latitude and longitude information and the Solar-viewing geometry are 450*270, but the reflectance data of each band in the EV_1KM_RefSB table is 1350*2029. How can I find the reflectance corresponding to the latitude and longitude and the Solar-viewing geometry?

Thanks.

The MOD021KM files contain radiances and reflectances, What's stored in the files are HDF Scientific Data Sets (SDSs), each of which has a "long_name" attribute that ends with "Scaled Integers". The values store in those SDSs are scaled integers, (abbreviated SI). You need to use these SI and the metadata stored in the file to convert to radiance using:
radiance(b) = radiance_scales*SI(b) - radiance_offsets(b)
So, you need to get scale and offset values from the original hdf file to calculate radiances.

For example:
The file MOD021KM.A2008297.0140.061.2017257054122.hdf has embedded metadata for EV_1KM_RefSB layer as below:
16-bit unsigned integer, 15 x 2030 x 1354
Number of attributes = 14
long_name = Earth View 1KM Reflective Solar Bands Scaled Integers
units = none
valid_range = 0,32767
_FillValue = 65535
band_names = 8,9,10,11,12,13lo,13hi,14lo,14hi,15,16,17,18,19,26
radiance_scales = 0.013213066,0.0076212585,0.0043131677,0.0031298073,0.0026987398,0.001104228,8.1663346E-4,0.0011990482,6.620522E-4,9.42601E-4,8.926971E-4,0.0070036417,0.008949267,0.006812885,0.0029916759
radiance_offsets = 316.9722,316.9722,316.9722,316.9722,316.9722,316.9722,316.9722,316.9722,316.9722,316.9722,316.9722,316.9722,316.9722,316.9722,316.9722
radiance_units = Watts/m^2/micrometer/steradian
reflectance_scales = 2.3534687E-5,1.2447986E-5,6.7703613E-6,5.1651145E-6,4.4347767E-6,2.2178097E-6,1.6401846E-6,2.4721908E-6,1.3650154E-6,2.263665E-6,2.8528423E-6,2.3296587E-5,3.1850235E-5,2.426077E-5,2.5489942E-5
reflectance_offsets = 316.9722,316.9722,316.9722,316.9722,316.9722,316.9722,316.9722,316.9722,316.9722,316.9722,316.9722,316.9722,316.9722,316.9722,316.9722
reflectance_units = none
corrected_counts_scales = 0.12619403,0.12619403,0.12619403,0.12619403,0.12619403,0.12619403,0.12619403,0.12619403,0.12619403,0.12619403,0.12619403,0.12619403,0.12619403,0.12619403,0.12619403
corrected_counts_offsets = 316.9722,316.9722,316.9722,316.9722,316.9722,316.9722,316.9722,316.9722,316.9722,316.9722,316.9722,316.9722,316.9722,316.9722,316.9722
corrected_counts_units = counts

Thank you very much for your reply! How do I correspond the reflectance read out to a specific longitude, latitude, and Solar-viewing geometry? Both are 450*270 matrices, while a certain band in EV_1KM_RefSB is 2023*1350. I want to obtain the longitude, latitude, and Solar-viewing geometry corresponding to the reflectance readout so as to compare it with other satellite data.
Many thanks in advance!

The MOD03 file has the lat/lon and view geometry at the same resolution as the L1B file. If these fields from an L1B file are to be used an interpolation to account for the different dimensions needs to be performed.