Has data changed recently?

[filoche]

Hi everyone.

I am back on a project I left in October 2023. This project involve extracting MODIS PAR data at several locations inside this bbox:

xmin ymin xmax ymax
-95.53711 57.65716 -58.71094 71.82884

As of today, I found out that the filenames to access the data has changed. For example, to download MONTHLY PAR of March-2003:

- Filename I was using in last October: A20030602003090.L3m_MO_PAR_par_4km.nc
- Filename as of today: AQUA_MODIS.20030301_20030331.L3m.MO.PAR.par.4km.nc

So I changed my code to adopt the new file naming convention.

However, I have noticed that the data also changed. On the attached plot, the top panel shows the data as it was in the file downloaded in October 2023, the bottom panel shows the data downloaded today (the circles represent the buffers in which I am extracting data).

Is it possible that the data have changed recently?

Thank you,
Phil

[OB ODPS - towens]

Yes, the files with the new names were reprocessed in Sept 2022 with the latest algorithms.

Tommy

[filoche]

Thank you for your reply.

With that new processing, averaging 8 years of monthly data (2003-2011, March-November), I am now getting 0 observations in my buffers until June/July.

Is that 'normal" that the new processing removed so much data? I mean, not a single value between March-June over 8 years.

[OB.DAACx - SeanBailey]

The PAR algorithm is only valid over ice-free waters. The R2022 reprocessing saw a number of changes to the ancillary inputs we use, one of these was the sea ice fraction. While the R2018 may have had some retrievals in your region of interest, they were likely not accurate. In the Hudson Bay, sea ice hangs around into June - or even later, so any PAR retrievals prior to June would be suspect.

Sean

[filoche]

Thank you, Sean, for the explanations.