Spectral profiles for different phases and custom products

[Wladimir Acevedo]

Hi dear community,

I'm working primarly with MTRDR products  for mineral identification using ENVI 5.3 + IDL with CAT 7.4 coupled and i have some questions: 

1.- Why when i visualize the spectral profile for some scene in any pixel the profile shape is essentially the same only with changes on the data value?. Does that mean there's only one dominant phase per scene?. Even when i use the Browse Products and can distinguish more than one phase (colors) the profile for the default image (true color) keeps its form. I don't know if this is a coincidence for the scenes that i've analyzed or there is something wrong on the procedure that i'm following.

2.- It is enough the interpretation with the continuum removed function or the ratioing process is necessary?. I consider that isn't an easy task find an appropiate denominator ROI. I've tried to emulate some examples from the MICA files but i did not get the same results. 

3.- About the custom stretching for Browse Products, when is neccesary fix the minimum value distinct to zero?. I understand that the maximum generally must be thw 99th percentile of the histogram, right?. 

4.-  How must be selected the right summary parameters  for the "Custom Products" when the "Browse Products" are not good enough?

 

Any help for any question will be very well received, thank you.

-Wladimir

[Feng]

Hi Wladimir,

 

Would you please provide the scenes you worked on so that we can better go through from there?

 

Thanks,

Feng

[Ray Arvidson]

Hi there: Would you please let me know which MTRDR scene you have been examining and I will take a look at the data and spectral plots and then answer your questions. 

Ray Arvidson

PDS Geosciences Node Manager and CRISM Science Team Member

[Wladimir Acevedo]

On 6/19/2019 at 8:47 AM, Ray Arvidson said:

Hi there: Would you please let me know which MTRDR scene you have been examining and I will take a look at the data and spectral plots and then answer your questions. 

Ray Arvidson

PDS Geosciences Node Manager and CRISM Science Team Member

Hi professor Arvidson,

The exact scene product is the: FRT000064D9_07_IF166J_MTR3. I've tried emulate the results from the MICA files example, but without success.

This image shows the ROI's location for the numerator (X:174, Y:529) and denominator (X:110, Y:529).

 

This second, the mean spectra for the numerator and denominator:

And this is the resulting spectra (smoothed) from the ratioing process, which is pretty different from the  MICA files example:

Any advice for the procedure or if you detect something wrong will be very helpful.

Thank you so much,

Wladimir

[Wladimir Acevedo]

On 6/17/2019 at 5:05 PM, Feng said:

Hi Wladimir,

 

Would you please provide the scenes you worked on so that we can better go through from there?

 

Thanks,

Feng

Hi Feng,

Some of the scenes that i've checked are:

FRT000064D9_07_IF166J_MTR3
FRT000064D9_07_IF166J_TER3 
FRT00009786_07_IF165J_MTR3
FRT00003E12_07_IF166J_MTR3
FRT00003E12_07_IF166J_TER3
FRT00008389_07_IF166J_MTR3
FRT000094F6_07_IF166J_MTR3
FRT0000634B_07_IF163J_MTR3
HRL000119C7_07_IF181L_TRR3

The big part of them were on the MICA files, i've tried to replicate the results from there as a starting work since i have not much expertise with this kind of data treatment.

Thank you,

Wladimir