For the site listed in the Table below, attached are spectral trajectory plots over time for Sentinel 2 spectral bands 2 (green) and 9 (SWIR2). The plots for bands 1, 2 and 4 are similar to 3;  and 8 is similar to 9.

In the plot for band 2, you’ll see distinctive jumps at date 5 (DoY 53, 22-02) and date 8 (DoY 88, 29-03).  This is also evident in the SWIR plots, though to a lesser degree. (There are arguably also blips for band 5 (RE-2) and for the NIR and RE-3 plots.)

The patterns seem to be consistent either side of the jumps, so whatever caused the jumps, it isn’t long-lasting.

The rainfall records for Fès Meknès Morocco show that there was rainfall of around 6 mm on 21^st / 22^nd February (and lower temperatures), and of around 15 mm on 15^th March (and lower temperatures). There was also rainfall of around 30 mm on 6^th March. At this stage, I don’t know whether the rain fell on the test plots.

I’m trying to understand how the rainfall may have affected the spectral response.

Fig 4 of Ogden et al “Removing Moisture Effect on Soil Reflectance Properties: A Case Study of Clay Content Prediction”, Fig 2 of Streck et al “Spectral signature of selected soils”, and several Google responses to “soil spectra for wet and dry soils”, show the spectral responses for wet soils to be less than those for dry soils. The attached plots show that the jumps are upwards – for the green band, from around 4.5%, to around 7.5% at date 5, and around 12% at date 8. However, as the papers show, wetting should lower the levels.

It seems to me that if rainfall accelerates the crop growth, then the change in spectral response should be sustained, and not drop back to the pre-jump levels.

If the rainfalls result in outbreaks of weeds, and these then die off, that could explain the jumps in spectral responses and the reversions to pre-jump levels.

Comments on the most likely reasons for the jumps in response would be appreciated.

APNI_plot_1_CSV_10m_LL_Morocco_34.06020_-4.76906_lines_41_pixels_41_times_12_bands_9_Fès-Meknès_SEFROU_Wheat_S2.csv

 

date :     1  2021-01-18   18  365

date :     2  2021-01-28   28  365

date :     3  2021-02-02   33  365

date :     4  2021-02-17   48  365

date :     5  2021-02-22   53  365

date :     6  2021-03-14   73  365

date :     7  2021-03-24   83  365

date :     8  2021-03-29   88  365

date :     9  2021-04-18  108  365

date :    10  2021-04-23  113  365

date :    11  2021-05-08  128  365

date :    12  2021-05-13  133  365