MCD19A2 data Processing

Use this Forum to find information on, or ask a question about, NASA Earth Science data.
Post Reply
xiaochuan_manchester
Posts: 5
Joined: Tue Mar 19, 2024 11:23 am America/New_York
Answers: 0

MCD19A2 data Processing

by xiaochuan_manchester » Mon Jan 06, 2025 11:31 am America/New_York

Hi, Sir:
I download the MCD19A2 Version 6 data with HDF format, while the dataset exclude the longitude and latitude information. Hence
1). I want to know how to match the each grid with location.
2). The Aerosol Optical Depth at 047 micron (or Aerosol Optical Depth at 055 micron) has different number of orbit overpasses, such as 3 x 1200 x 1200 or 5 x 1200 x 1200 or 6 x 1200 x 1200 in different date at the same region, I am not sure which the number of orbit overpasses should be selected.

Filters:

LP DAACx - lien
Posts: 248
Joined: Thu Jun 25, 2020 9:51 am America/New_York
Answers: 2
Been thanked: 9 times

Re: MCD19A2 data Processing

by LP DAACx - lien » Mon Jan 06, 2025 1:30 pm America/New_York

Hello,
This site has links to the tile Gring and Bounding coordinates: https://modis-land.gsfc.nasa.gov/MODLAND_grid.html
Please note the Gring and Bounding coordinates are not exact. This script will give you the exact
coordinates of any specific pixel within a given MODIS tile: https://lpdaacsvc.cr.usgs.gov/services/tilemap?help=html

Please be aware that the MODIS sensors have a nadir revisiting cycle of every16 days. At and near the equator the earth is not collected every day, so there are data gaps. Farther away from the equator areas are collected multiple times a day. However, the pixels on any given tile are not necessarily collected from the same swath. So that is why you see the variance from day to day. Any given pixel will not necessarily have been covered the same number of times from one day to the next. One thing for sure though is that every16 days any given pixel will have the same number of overpasses at the same viewing angle.
Thanks,
Brett

xiaochuan_manchester
Posts: 5
Joined: Tue Mar 19, 2024 11:23 am America/New_York
Answers: 0

Re: MCD19A2 data Processing

by xiaochuan_manchester » Tue Jan 07, 2025 8:56 am America/New_York

Hi, Thanks for your help.

I open the link: https://lpdaacsvc.cr.usgs.gov/services/tilemap?help=html, while it shows below figure. I still indeed have no idea to figure out how to correspondence between the pixel and the coordinate in MCD19A2 file? I am not sure if the link is wrong or cannot open correctly.

I have several known points with their longitude and latitude, I want to extract the pixel information in MCD19A2 file. I need to locate to a specific pixel location according to the known point.

Please help me solve it, Thanks a lot.
Attachments
the figure of https://lpdaacsvc.cr.usgs.gov/services/tilemap?help=html
the figure of https://lpdaacsvc.cr.usgs.gov/services/tilemap?help=html
Fitu.JPG (123.82 KiB) Not viewed yet

xiaochuan_manchester
Posts: 5
Joined: Tue Mar 19, 2024 11:23 am America/New_York
Answers: 0

Re: MCD19A2 data Processing

by xiaochuan_manchester » Tue Jan 14, 2025 11:35 am America/New_York

Hi, Sir:

Considering the absence of geographic information layer in MCD019A2, which unable to locate the specific point with known longitude and latitude. Could you please give me some suggestion about how to locate it. Furthermore, due to there are different layers in the same sds (Optical_Depth_055), I am not sure which layer is best available.

Thanks a lot

hyoklee
Posts: 12
Joined: Fri Mar 20, 2020 9:24 pm America/New_York
Answers: 0
Has thanked: 2 times

Re: MCD19A2 data Processing

by hyoklee » Tue Jan 14, 2025 11:50 am America/New_York


xiaochuan_manchester
Posts: 5
Joined: Tue Mar 19, 2024 11:23 am America/New_York
Answers: 0

Re: MCD19A2 data Processing

by xiaochuan_manchester » Tue Jan 14, 2025 12:18 pm America/New_York

hyoklee wrote:
> This Python code may help you:
>
>
> https://www.hdfeos.org/zoo/MORE/LPDAAC/MCD/MCD19A2.A2010010.h25v06.006.2018047103710.hdf.py

Hi, sir

Thanks for your help.
Indeed, I have downloaded a lot of MCD19A2 file (version 6.1) with HDF4 format, while I am failed to how to extract the available AOD in these files due to I don't know the geographic information.

for example, like following file: MCD19A2.A2002279.h17v07.061.2022203022902.hdf. This file cover the Agoufou site (lon: -1.479117, lat: 15.3454) and I want to extract the AOD value.
1).Because of I don't know the exact geographic information (longitude and latitude information about this file), I cannot calculate the specific grid location after obtain the SDS of Optical_Depth_055 or other SDS.
2). For instance, four layers in the SDS of Optical_Depth_055, I am not sure which layer or best layers for extraction.
Attachments
MCD19A2.A2002279.h17v07.061.2022203022902
MCD19A2.A2002279.h17v07.061.2022203022902
MCD19A2.JPG (138.54 KiB) Not viewed yet

hyoklee
Posts: 12
Joined: Fri Mar 20, 2020 9:24 pm America/New_York
Answers: 0
Has thanked: 2 times

Re: MCD19A2 data Processing

by hyoklee » Tue Jan 14, 2025 11:37 pm America/New_York

Here's another example that subsets data at specific location:

https://hdfeos.org/zoo/MORE/LPDAAC/MCD/MCD19A2.A2010010.h25v06.006.2018047103710.hdf.s.py

Please adjust the following code for your region:

Code: Select all

# Subset region.
# lon = 85 : 90 E
# lat = 20 : 25 N

latbounds = [ 20 , 25 ]
lonbounds = [ 85 , 90 ]
According to [1], the 4 layers correspond to orbits.
Thus, I think you can choose any orbit that doesn't have fill value at your location.
Or you can choose one based on QA dataset, I guess.

[1] https://lpdaac.usgs.gov/products/mcd19a2v061/#:~:text=The%20MCD19A2%20AOD%20data%20product,composite%20of%20all%20available%20orbits.

xiaochuan_manchester
Posts: 5
Joined: Tue Mar 19, 2024 11:23 am America/New_York
Answers: 0

Re: MCD19A2 data Processing

by xiaochuan_manchester » Wed Jan 15, 2025 1:56 pm America/New_York

Hi, Sir

Thank you very much for your help.

However, I still cannot resolve my original problem. As shown in your supported link, it can visualize regional AOD. Indeed, I don't need to visualize this step firstly, I just want to extract the AOD information at specific pixel according to the known point's longitude and latitude. Due to referred amounts of downloaded data in HDF format so that I cannot convert each HDF file into TIFF and then extract the AOD information.

Before, I can extract the information (MOD021KM) at specific pixel according to the MOD03 file, while no geographic information in MCD19A2.
I have researched a lot of materials in this part, but I still cannot resolve it. It is much let us depressed.

Below from LP DAAC - lien while it's not exact as he said:
This site has links to the tile Gring and Bounding coordinates: https://modis-land.gsfc.nasa.gov/MODLAND_grid.html
Please note the Gring and Bounding coordinates are not exact. This script will give you the exact
coordinates of any specific pixel within a given MODIS tile: https://lpdaacsvc.cr.usgs.gov/services/tilemap?help=html

LP DAAC - dgolon
User Services
User Services
Posts: 44
Joined: Tue Dec 03, 2024 2:37 pm America/New_York
Answers: 0
Has thanked: 11 times

Re: MCD19A2 data Processing

by LP DAAC - dgolon » Tue Jan 28, 2025 10:57 am America/New_York

Just a quick note for the thread. While we do not have a specific answer to xiaochuan_manchester's question, and we give tremendous thanks to the HDF Group for all their work to help our users, we do have a minor update on the tool. The LDOPE team is aware of the issue, and it is on their backlog to address in the future. However, due to limited resources we do not have an ETA on when the issue with this portion of the tool will be resolved.
Subscribe to the LP DAAC listserv by sending a blank email to lpdaac-join@lists.nasa.gov.

Sign up for the Landsat listserv to receive the most up to date information about Landsat data: https://public.govdelivery.com/accounts/USDOIGS/subscriber/new#tab1.

Post Reply