How to make a time series based on pixel location

[daviluca]

How to make a time series based on pixel location
I have had a lot of difficulty keeping track of what the LST pixel value is since every 8 days image the pixel ID changes. What I'm trying to do is know the pixel temperature value over the weeks. Is this possible? I really need help a lot.
I am using AppEEARS.
Thank you very much for any help.
David Luca

[LP DAAC - jwilson]

We have forwarded your inquiry to a Subject Matter Expert for further assistance.

[LP DAAC - afriesz]

Good morning @daviluca ,

The day of year (doy) for each geoTIFF output from AppEEARS can be found in the file name. For example, the doy for this AppEEARS output, MOD11A2.006_LST_Day_1km_doy2021129_aid0001.tif, is 129 for the year 2021 (doy2021129). This translates to the date 05/09/2021 (https://www.esrl.noaa.gov/gmd/grad/neubrew/Calendar.jsp).

Regarding the data values. AppEEARS outputs for Area samples are in their original packed form, or rather, they have not had the scale factor and offset applied. In the case of the MODIS Land Surface Temperature science dataset in MOD11A2 version 6, the scale factor is 0.02 with an offset of 0. One applies the scale factor and offset using the follow equation:

data_value = (packed_value * scale_factor) + add_offset

A real example looks like this: 301.92 = (15096 * 0.02) + 0, where 301.92 is the ‘unpacked’ value in Kelvin and 15096 is the data value from the source data file/AppEEARS Area sample output. Applying the scale factor like this to a raster file can be done using the ‘Raster Calculator’ in ArcGIS or in QGIS.

For MODIS Land Surface Temperature, the units for the resulting data value is in Kelvin (K). You can then make the conversion to Celsius or Fahrenheit once you have applied the scale factor.

[daviluca]

Thank you very much for the helpful reply. But there is little to get where I intend. I hit the src of the image, I observed the image date, but the value I need from the 1km pixel. I did what was recommended but I have a slight deviation in the image, but if there is a way to make such correction it would be really perfect. I put the images to demonstrate.

[daviluca]

Here I show on the map that when I overlay the images the geolocation of the pixel has different IDs. Same day but different years. 2019 to 2021.

[LP DAAC - afriesz]

@daviluca ,

Would you be able to share the vector files you're using and your AppEEARS request ID? If so, please send to LPDAAC@usgs.gov with the subject "Forum: How to make a time series based on pixel location".