Skip to end of metadata
Go to start of metadata




Grant Agreement

776276

Acronym

PLANMAP

Project full title

Planetary mapping



Deliverable

D 4.4

Deliverable Name

Retrieved compositional units

Nature of deliverable


Dissemination level

PU

Scheduled delivery date`

30th Sep 2020

Status




Prepared by:

Francesca Zambon, Cristian Carli, Francesca Altieri, Jean-Philippe Combe

Verified by:

Sabrina Ferrari, Jack Wright

Approved by:

Matteo Massironi




List of Acronyms

AcronymDescription
VNIRvisible-near infrared
PCprincipal component
M3Moon Mineralogy Mapper
CRISMCompact Reconnaissance Imaging Spectrometer for Mars
MROMars Reconnaissance Orbiter
MESSENGERMErcury Surface, Space ENvironment, GEochemistry, and Ranging
TESThermal Emission Spectrometer
THEMISThermal Emission Imaging System
SUSpectral Unit
SISpectral slope I
SIISpectral slope II
GSGlobal Slope
R750Reflectance at 750 nm
R540Reflectance at 540 nm
BDBand Depth
SSBISpectral Slope Band I

Executive summary


Terrestrial geological maps are made by combining not only morphology and stratigraphy, but also mineralogical and/or compositional information (e.g. Bernknopf et al., 1993).

For Earth, unlike the other bodies of our Solar System, it is easy to study bedrock in situ and to take samples to analyse later in a laboratory, allowing for a detailed petrographic and mineralogical description of the samples. Fieldwork, integrated with remote sensing observations, is the best way to produce comprehensive geological maps.

Although in the last decades several space missions equipped with instruments dedicated to the study of the surface composition explored many planetary bodies collecting a huge amount of data, it is not still possible to produce geological maps comparable to Earth ones.

The main issues are due mainly to a lack of field samples, incomplete datasets, and/or different spatial resolutions, sometimes too low for extracting the necessary information, or acquisition at bad illumination geometries, inducing photometric artefacts.

Despite the difficulties, it is still possible in some cases to retrieve and integrate morpho-stratigraphic and mineralogical and/or compositional maps, obtaining a final product similar to the terrestrial geological maps.

In this project, we consider visible-near infrared (VNIR) multi- and hyperspectral data of selected regions of Mercury, the Moon, and Mars. To define the spectral units (SUs), we retrieved and applied for the first time, when not available from the literature, proper spectral parameters based on the spectral characteristics of the target areas (see D4.2, D4.3).

We determined spectral indices that highlight the spectral variability of different planetary surfaces. Some of them are strictly dependent on the regions considered (e.g. the principal component analysis (PCA) applied to highlight spectral variability within the Hokusai quadrangle of Mercury) while others, such as spectral slopes and band depths, have a global application and are independent of the analysed region. 

SUs are derived by considering all the spectral parameters selected for a given region together. Each unit defines the mineralogical and/or compositional characteristics of a given region, which sometimes could be associated with morphological variations (See next deliverable D3.4 - Update of the stratigraphic chart after the integration with compositional units).

In some cases, when clear absorption bands are evident, a SU can indicate a mineralogical identification (e.g. pyroxenes on the Moon). Conversely, if the spectra do not show clear absoption bands (e.g. in Mercury's case), we are not able to identify a specific mineralogical phase, but rather we observe only spectral slope variations.

The final WP4 deliverable, D4.4, concerns the release of the spectral units maps for the Hokusai quadrangle on Mercury, the Apollo basin and Von Karman/Leibnitz crater regions on the Moon, and Crommelin crater and Arsinoes Chaos on Mars. The products released in this deliverable are innovative because, unlike the maps currently available in the literature, we summarise in a single map all the information held in several spectral indices maps, allowing for an easy and quick understanding of the mineralogical and compositional properties of each region. This work required the collaboration and the expertise of different PLANAMP WPs, in particular WP2, WP3, and WP5. This collaboration will continue until the end of the project, including WP7 data fusion support. Furthermore     , some of these products will be used by WP6 for 3D model production and by WP5 for      virtual reality integration. D4.4 confirms the importance of the interaction among the various WPs in past months and in the final part of the project.

This approach plays a strategic role as a forerunner for future projects, integrating information related to mineralogy from thermal infrared, or elemental information from X-ray datasets, where present at scale similar to those reported by VNIR imaging instruments, resulting in a more complete compositional picture for planetary surfaces.

In the next sections we describe in detail the method used to define the SUs for the regions of interest selected for Mercury, the Moon, and Mars, showing the final spectral units obtained for each case.

Spectral Units Definition Approach


In this section we show the final products released by WP4. The SUs presented here have been obtained from the spectral parameters delivered in D4.2 and D4.3. For further information on the spectral indices selection and retrieval, we refer to D4.2 and D4.3.

Mercury


In this section we focus on the SU definition for the Hokusai quadrangle of Mercury, which is one of the main targets of the PLANMAP project. The morpho-stratigraphic map of this region has been published by Wright et al. 2019 and will be integrated with these SU maps in subsequent deliverables.

Hokusai quadrangle, located in Mercury's northern hemisphere, is particularly interesting and spectrally diverse, due to the presence of different terrains; e.g. the northern smooth plains of Borealis Planitia, cratered terrains, dark and bright material, as well as a series of faculae (diffuse, bright, reddish deposits) associated with putative explosive volcanic vents. The most relevant example is Nathair Facula, the brightest and most prominent facula on the planet (Rothery et al., 2020, submitted). Furthermore, Hokusai quadrangle includes several craters with fresh ejecta including the Hokusai      crater, whose name is given to the quadrangle, and whose ray system (the longest on Mercury) extends over much of the planet. 

Here we consider the spectral parameters released in D4.2 and D4.3, in particular, the reflectance at 750 nm, and three spectral slopes, SI, SII, and GS, calculated respectively between 430 and 560 nm, 750 and 1000 nm and 430 and 1000 nm (see D4.2).

Below, we have summarised the steps of the method applied to define the SUs: 

  • We considered the histograms of R750, SI, SII and GS. We divided the range of values for each spectral parameter into a fixed number of intervals (seven for Mercury) based on their histogram values distribution (see Fig. 1) and on their density scatter plots (Figs 2 and 3), then we chose the threshold values considering the mode of the distribution as a reference value.

 We named the thresholds intervals, starting from the lower to the higher range of values, as follows:

  • Very Low values (VL)
  • Low values (L),
  • Intermediate Low values (IL),
  • Intermediate values (I),
  • Intermediate High values (IH),
  • High values (H),
  • Very High values (VH).


The threshold values for each parameter are reported in the tables of Figs 4–7.

2) We assigned to each threshold interval a progressive number, VL= 1, L=2, IL=3, I=4, IH=5 H=6, VH=7, and we found all the possible combinations of these intervals for all the parameters. 

3) The last step consists of the selection of the final spectral units. We selected all the most populated units and we merged together all the units with similar spectral characteristics, excluding those sparsely populated and less significant, while considering the final map resolution.


Figure 1: Hokusai histograms for R750 (panel a), SI (panel b), GS (panel c), SII (panel d). Green lines indicate the threshold limits, while the blue line refers to the mode of the distribution.


Figure 2: GS vs R750 density scatter plot and associated histograms as shown in Fig. 1. Red lines indicate the interval thresholds, while the blue line indicates the mode value.

Figure 3: SI vs SII density scatter plot and associated histograms as shown in Fig. 1. Red lines indicate the interval thresholds, while the blue line indicates the mode value.


Parameter

R750

Red (VL)

Green (L)

Sea Green (IL)

Blue (I)

Cyan (IH)

Magenta (H)

Coral (VH)

Mode =  0.08

< 0.0560.056–0.0650.065–0.0750.075–0.0880.088–0.0970.097–0.123>0.123

#pixels

(24550704)

673392

(2.7 %)

2651183

(10.8 %)

6300790

(25.7 %)

9545046

(38.9 %)

3336032

(13.6 %)

1796124

(7.3 %)

248137 

(1.0 %)

Figure 4: Map in panel a) displays the thresholds classes obtained for R750. Each colour refers to a specific class. Threshold intervals are reported in the bottom table. Panel b) shows the average reflectance spectra of the corresponding classes, while panel c) exhibits the same spectra normalised at 560 nm.

Parameter

SI

Red (VL)


Green (L)


Sea Green (IL)


Blue (I)


Cyan (IH)


Magenta (H)


Coral (VH)


Mode = 2.66 1/µm

<2.2 1/µm2.2–2.39 1/µm2.39–2.49 1/µm2.49–2.81 1/µm2.81–2.94 1/µm2.94–3.15 1/µm>3.15 1/µm

#pixels

(24550704)

1802484

(7.3 %)

2543455 

(10.4 %)

2319655

(9.4 %)

9957066

(40.6 %)

3117022

(12.7 %)

2762398 

(11.3 %)

2048624

(8.3 %)


Figure 5: Panel a) shows the threshold classes obtained for SI. Panels b) and c) show the average reflectance spectra of the corresponding classes before and after the normalisation at 560 nm respectively.  


Parameter

SII  

Red (VL)


Green (L)


Sea Green (IL)


Blue (I)


Cyan (IH)


Magenta (H)


Coral (VH)


Mode = 0.902 1/µm

<0.64 1/µm0.64–0.72 1/µm0.72–0.79 1/µm0.79–1.08 1/µm1.00 1/µm1.08 1/µm1.18 1/µm

#pixels

(24550704)

1139393

(4.6 %)

1540170

(6.3 %)

2579147

(10.5 %)

11534180

(47.0 %)

3432817

(14.0 %)

2484584

(10.1 %)

1840413

(7.5 %)

Figure 6: Map in panel a) displays the threshold classes obtained for SII. As with Figs 4 and 5, each colour refers to a specific class and the threshold values for each interval are reported in the table underneath. Panel b) shows the average reflectance spectra of the corresponding classes, while panel c) shows the same spectra normalised at 560 nm.

Parameter

GS

Red (VL)


Green (L)


Sea Green (IL)


Blue (I)


Cyan (IH)


Magenta (H)


Coral (VH)


Mode = 1.86 1/µm

< 1.49 1/µm1.49–1.70 1/µm1.70–1.94 1/µm1.94–2.2 1/µm2.2–2.4 1/µm2.4–2.8 1/µm>2.8 1/µm

#pixels

(24550704)

949550

(3.9 %)

4051894

(16.5 %)

8130707

(33.1 %)

7210096

(29.4 %)

2928639

(11.9 %)

1132971

(4.6 %)

146847

(0.6 %)

    

Figure 7:  Threshold values for GS (panel a). In panels b) and c) the average reflectance spectra of the corresponding classes are reported, before and after the normalisation at 560 nm respectively. The bottom table shows the threshold values for each interval for GS.

  

Figure 8: Hokusai quadrangle spectral units map obtained by the spectral thresholding shown in Figs 4–7 (panel a). Panel b) shows the average reflectance spectra of Hokusai quadrangle SUs, while in panel b) are reported the same spectra normalised at 560 nm to better emphasise spectral variations not associated with the reflectance. Black areas represent unclassified units.

Hokusai spectral units definition:

The following table summarises the spectral units obtained for the Hokusai quadrangle.

Table 1

Spectral UnitProposed nameSpectral indicationSpectral Parameters RangeAssociated Morphostratigraphic Unit

#1 

Purple

Very Bright Materials (VBM)Very high R750 and very shallow SII and GS spectral slopes

R750: H–VH

SI: IL–I

SII: VL–L

GS: VL–L

Crater ejecta, crater wall, bright spots, very fresh material

#2

Sea green

Intermediate Bright Materials (IBM)High R750 and shallow spectral slopes

R750: IH–H

SI: L–I

SII: VL–IL

GS: L–IL

Crater rays, crater ejecta, bright spots, fresh material

#3

Blue

Low Bright Materials (LBM)bright material with shallower spectral slopes than the Hokusai average, but lower than units #1, #2

R750: IH–H

SI: L–I

SII: VL–L

GS: VL–L

Degraded rays and ejecta

#4

Red

Rachmaninoff Dark Materials (RDM)Very low R750, very shallow SI and GS, shallow SII

R750: VL–L

SI: VL–L

SII: VL–IL

GS: VL–L

Floor of Rachmaninoff crater

#5

Green

Intermediate Units (IU)Intermediate values of the spectral parameters

R750: VL–IL

SI: I–IH

SII: I–H

GS: I–H

Cratered terrains, intermediate terrains

#6

Cyan

High Reflectance and spectral Slopes Units (HRSU)High reflectance and steep spectral slopes

R750: IL–IH

SI: IH–VH

SII: I–H

GS: I–H

Faculae and Borealis Planitia

#7

Coral

Nathair Facula (Nat)Very high reflectance and very steep spectral slopes

R750: H–VH

SI: H–VH

SII: I–IH

GS: I–IH

Nathair Facula

#8

Magenta

Intermediate dark material (IDM)Intermediate low R750, shallow spectral slopes

R750: IL–I

SI: L–I

SII: L–I

GS: VL–IL

Intermediate terrains

#9

Pink

Intermediate material low spectral slopes (IMLS)Intermediate–high R750 and intermediate low SI and GS

R750: I–H

SI: IL–IH

SII: L–I

GS: L–I

Hokusai ejecta

#10

Dark Yellow

Dark material low spectral slope (DMLS)Very low–low R750, shallow spectral slopes

R750: VL–L

SI: L–I

SII: L–I

GS: L–I

Low reflectance intermediate terrains

#11

Dark Blue

Low reflectance material, high spectral slopes (LRHS)Low R750, intermediate high–very high spectral slopes

R750: VL–IL

SI: IH–VH

SII: IH–VH

GS: IH–VH

Borealis Planitia

#12 

Black

Unclassified Pixels unit (UP)All the regions not included in the previous units. Mainly regions affected by photometric artefacts.



Moon


Here we derived the SUs of two regions of interest already released in D4.3: the Apollo basin, and the Von Karman/Leibnitz craters region. 

These regions, located within maria on the lunar farside, are characterised by material of various ages. Furthermore, specific smooth units are highlighted (e.g. Ivanov et al. 2018). Since the regions considered are quite large and the available M3 data have a coverage and spatial resolution suitable for the creation of mosaics of the entire regions, we used an approach similar to the Mercury one to derive Apollo and Von Karmar/Leibnitz region SUs.

Due to the unavailability of public high-level spectral products, a global analysis of the Moon's spectral parameters is not possible at this stage. This topic will be the goal of future projects such as GMAP within the EuroPlanet 20-24.

For the Moon case we followed a twofold approach to define SUs. First, we calculated the SUs of Apollo basin and of Von Karman/Leibnitz craters separately, considering them as two distinct regions. We defined the SUs from the spectral parameters delivered in D4.3, in particular: reflectance at 540 nm (R540), band depths at 1 µm (BDI) and 2 µm (BDII) and the spectral slope between 540 nm and the maximum of the second shoulder of the band depth at 1 µm (SSBI) (for more details on the spectral parameters definition see D4.2 and D4.3). This approach allows us to define local spectral units specific for each region of interest.

Second, we define SUs for the Moon by merging the Apollo basin and Von Karman/Leibnitz craters into one map and then by considering a single spectral parameter histogram for both regions, obtaining a common SUs definition.

Although we found similar SUs definitions, the SUs ranges of values are different when we consider the Apollo      basin and the Von Karman and Leibnitz craters regions together or separately (see section 1 and 2). 

For the Moon in both the cases, we used a method similar to those applied for Mercury to derive the SUs, with a lower number of spectral parameter threshold classes.

For each spectral parameter (see D4.2, D4.3), we considered the histogram values and the scatter plots shown in Fig. 9, then we defined five threshold values classes always considering the mode of the distribution and the spectral parameters density scatter plots (Figs 10–11). 

The threshold intervals are the following:

  • Low values (L),
  • Intermediate Low values (IL),
  • Intermediate values (I),
  • High values (H),
  • High values (IH).

The range of each threshold interval varies according to the spectral parameters values distribution. The threshold values of the Moon parameters are shown in Figs 12–15, 20–23, 28–31.

We define the SUs as indicated in the steps 2 and 3 of Mercury section.

Spectral and morphological units integration will be discussed in D3.4. 

Case 1: Apollo Basin & Von Karman and Leibnitz craters studied separately


In this section we report the Moon results obtained calculating the spectral units considering Apollo and Von Karman region as two distinct areas.

Apollo basin


Figure 9: Histograms of the spectral parameters values retrieved in D4.3: a) reflectance at 540 nm, b) band depth at 1 µm, c) spectral slope between the maximum value of the band at 1 µm shoulders. The blue vertical line represents the mode value of the distribution, while the green lines show the threshold values.

Figure 10: Density scatter plot of the band depth at 2 µm (BDII) vs band depth at 1 µm (BDI). Blue line indicates the mode value for both the parameters and red lines within the plot show the threshold values. For clarity, we reproduce the corresponding histograms shown in Fig 9b and d.



Figure 11: Density scatter plot of the spectral slope relative to the band at 1 µm (SSBI) vs reflectance at 540 (R540), together with the corresponding histogram values. Threshold lines and mode lines      follow the same notation of Figs 9–10.




Parameter

R540

Red (L)

Green (IL)

Blue (I)

Cyan (IH)

Magenta (H)

Mode = 0.044

<0.029

0.029–0.037

0.037–0.051

0.051–0.065

>0.065

#pixels

(1371514)

42873

(3.1 %)

87844

(6.4 %)

715442

(52.2 %)

433433

(31.6 %)

91922

(6.7 %)

   

Figure 12: R540 map of Apollo basin, based on the intervals shown in Figures 10 and 11 (panel a). We indicate in red low (L) reflectance values, in green intermediate low (IL) values, in blue intermediate (I) values, in cyan intermediate high (IH) reflectance values and in magenta high reflectance values (H). Panel b) plot shows the average reflectance spectra of the same colour classes displayed in panel a). Panel b) plot represents the same spectra of b) normalised at 540 nm, while plot d) show the spectra of panel b) after continuum removal. For the continuum removal details, and the reflectance map retrieval see D4.3. The bottom table reports the threshold values of the reflectance for each threshold class.



Parameter

SSBI

Red (L)


Green (IL)


Blue (I)


Cyan (IH)


Magenta (H)

Mode = 1.55 1/µm

<1.35 1/µm

1.35–1.48 1/µm

1.48–1.66 1/µm

1.66–1.73 1/µm

>1.73 1/µm

#pixels

(1370097)

178002

(13.0 %)

369310

(27.0 %)

703871

(51.4 %)

95384

(7.0 %)

23530

(1.7 %)

    

Figure 13:  SSBI map of Apollo basin (panel a). Colour scheme for the intervals is the same as shown in Fig. 12. Plot in panel b) shows average reflectance spectra using the same colour scheme displayed in panel a). Panel c) plot shows the spectra of b) normalised at 540 nm, while plot d) shows the spectra of b) after continuum removal. The bottom table reports the threshold values of the reflectance for each class.



Parameter

BDI

Red (L)

Green (IL)

Blue (I)

Cyan (IH)

Magenta (H)

Mode = 0.049

<0.027

0.027–0.038

0.038–0.063

0.063–0.088

>0.088

#pixels

(1370928)

38828

(2.8 %)

132293

(9.6 %)

615668

(44.9 %)

380253

(27.7 %)

203886

(14.9 %)


Figure 14: BDI map of the Apollo basin (panel a). Colour scheme for the classes is the same as shown in Fig. 12 and 13. Plot in panel b) shows average reflectance spectra using the same colour scheme displayed in panel a). Panel c) plot represents the spectra of b) normalised at 540 nm, while plot d) shows the spectra of panel b) after continuum removal. The bottom table reports the threshold values of the reflectance for each class.



Parameter

BDII

Red (L)

Green (IL)

Blue (I)

Cyan (IH)

Magenta (H)

Mode = 0.056

<0.04

0.04–0.048

0.048–0.063

0.063–0.078

>0.078

#pixels

(1373577)

34448

(2.5 %)

145916

(10.6 %)

630702

(45.9 %)

378103

(27.5 %)

184408

(13.4 %)


   

Figure 15: BDII thresholds map of Apollo basin (panel a). Colour scheme for the intervals threshold classes is the same shown in Fig. 12, 13 and 14. Plot in panel b) shows average reflectance spectra using the same colour scheme displayed in panel a). Panel c) represents the spectra of b) normalised at 540 nm, while plot d) shows the spectra of panel b) after continuum removal. The bottom table reports the threshold values of the reflectance for each class.



Figure 16: Spectral units map of Apollo basin based on the threshold intervals shown in the previous figure (panel a). Black areas are unclassified units. Plot in panel b) shows average reflectance spectra using the same colour scheme displayed in panel a). Panel c) plot represents the same spectra of b) normalised at 540 nm, while plot d) shows the spectra of panel b) after the continuum removal. 

Von Karman/Leibnitz area

Here we show the SUs obtained applying the same method used for the Apollo region to Von Karman/Leibnitz area.

Figure 17: Histogram values for lunar spectral parameters calculated for Von Karman/Leibnitz area. The threshold values lines are green, while the mode line is blue.

Figure 18: SSBI vs R540 scatter plot for the Von Karman/Leibnitz area. The threshold values within the scatter plot are red while the mode line is blue.

Figure 19: BDI vs BDII scatter plot for the Von Karman/Leibnitz area. The threshold and mode lines follow the same notation of Fig. 18.


Parameter

R540

Red (L)

Green (IL)

Blue (I)

Cyan (IH)

Magenta (H)

Mode =  0.045

<0.0320.032–0.0380.038–0.0530.053–0.06>0.06

#pixels

(885860)

22165

(2.5 %)

107934

(12.2 %)

567869

(64.1 %)

118357

(13.4 %)

69535

(7.8 %)

    

Figure 20: Panel a) shows R540 map for the Von Karman/Leibnitz area. Panel b) displays the reflectance average spectra of the threshold classes shown in panel a). Plots in panels c) and d) report the same spectra of panel b) normalised at 540 nm, and continuum-removed spectra to emphasise the spectral slopes, and band depths variation respectively.


     

Parameter

SSBI

Red (L)

Green (IL)

Blue (I)

Cyan (IH)

Magenta (H)

Mode = 1.49 1/µm

<1.32 1/µm1.32–1.41 1/µm1.41–1.58 1/µm1.58–1.66 1/µm> 1.66 1/µm

#pixels

(886112)

103515

(11.7 %)

149672

(16.9 %)

491275

(55.4 %)

111278

(12.6 %)

30372

(3.4 %)


Figure 21: Panel a) SSBI classes for the Von Karman/Leibnitz area. Panel b) displays the average reflectance spectra of the threshold classes shown in panel a). Plots in panels c) and d) report the same spectra of panel b) normalised at 540 nm, and continuum-removed spectra to emphasise the spectral slopes, and band depths variations respectively.



Parameter

BDI

Red (L)

Green (IL)

Blue (I)

Cyan (IH)

Magenta (H)

Mode = 0.086

<0.0510.051–0.0680.068–0.0980.098–0.115>1.115

#pixels

(884456)

26051

(2.9 %)

117573

(13.3 %)

476542

(53.9 %)

165295

(18.7 %)

98995

(11.2%)


Figure 22: Panel a) BDI threshold classes for the Von Karman/Leibnitz area. Panel b) displays the average reflectance spectra of the classes shown in panel a). Plots in panels c) and d) report the same spectra of panel b) normalised at 540 nm, and continuum-removed spectra to emphasise the spectral slopes, and band depths variations respectively.


   

Parameter

BDII

Red (L)

Green (IL)

Blue (I)

Cyan (IH)

Magenta (H)

Mode = 0.071

<0.0550.055–0.0630.063–0.0810.081–0.091>0.091

#pixels

(886909)

27414

(3.1 %)

101703

(11.5 %)

484948

(54.7 %)

151832

(17.1 %)

121012

(13.6%)


Figure 23: Panel a) shows BDII threshold classes for the Von Karman/Leibnitz area. Panel b) displays the average reflectance spectra of the threshold classes shown in panel a). Plots in panels c) and d) report the same spectra of panel b) normalised at 540 nm, and continuum-removed spectra to emphasise the spectral slopes, and band depths variations respectively.


Figure 24: Panel a) displays the spectral units map for the Von Karman/Leibnitz area, obtained using the spectral parameters reported in Figs 20–23. Panel b) shows the average reflectance spectra of the threshold classes shown in panel a). Plots in panels c) and d) report the same spectra of panel b) normalised at 540 nm, and continuum-removed spectra to emphasise the spectral slopes, and band depths variations respectively. Black areas represent unclassified units.

Apollo basin and Von Karman/Leibnitz spectral units definition: Case 1

Apollo basin and Von Karman/Leibnitz regions have in some case common SUs. Here we report in a single table the SUs for both the regions specifying where each unit is present. 

Although the SUs are the same for two areas considered, since the threshold classes have been defined for each region      independently, the spectral parameters ranges vary between the two areas, as shown in tables of Figs 12–15 and Figs 20–23.


Table 2

Spectral Unit

Location

Spectral Indication

Spectral Parameters Range

Associated Morphostratigraphic Unit

#1 (Red)

Mainly in Apollo region, few pixels in Von Karman/Leibnitz

Low reflectance values, from low to intermediate low spectral slope values, from intermediate high to high band depths values.

R540: L

SSBI: L–IL

BDI: IH–H

BDII: IH–H

Dark smooth plains 
#2 (Green)

Mainly in Apollo region, few pixels in Von Karman/Leibnitz

Intermediate low reflectance values, from low to intermediate low spectral slope values, from intermediate high to high band depths values.

R540: IL

SSBI: L–IL

BDI: IH–H

BDII: IH–H

Intermediate dark smooth plains
#3 (Cyan)Apollo, Von Karman/LeibnitzFrom intermediate low to intermediate reflectance values, from intermediate to intermediate high spectral slope values and from  intermediate high  to high band depths values.

R540: IL–I

SSBI: I–IH

BDI: IH–H

BDII: IH–H 

No evident correlation. To be discussed during the data integration.
#4 (Blue)Apollo, Von Karman/LeibnitzFrom intermediate low to intermediate reflectance, spectral slope and band depths values.

R540: IL–I

SSBI: IL–I

BDI: IL–I

BDII: IL–I

Intermediate cratered terrains
#5 (Magenta)Apollo, Von Karman/LeibnitzFrom intermediate high to high reflectance values, from low to intermediate low spectral slope values, from intermediate high to high band depths values.

R540: IH–H

SSBI: L–IL

BDI: IH–H

BDII: IH–H

Bright ejecta
#6 (Maroon)

Mainly in Apollo region, few pixels in Von Karman/Leibnitz

From intermediate high to high reflectance values, from intermediate to intermediate high spectral slope values and band depths values.

R540: IH–H

SSBI: I–IH

BDI: I–IH

BDII: I–IH

No evident correlation. To be discussed during the data integration.
#7 (Sea Green)Apollo, Von Karman/LeibnitzIntermediate reflectance and spectral slope values, from intermediate high to high values and band depths values.

R540: I

SSBI: I

BDI: IH–H

BDII: IH–H

Intermediate reflectance ejecta and terrains 
#8 (Purple) 

Mainly in Apollo region, few pixels in Von Karman/Leibnitz

From intermediate high to high reflectance values, from intermediate to intermediate high spectral slope values, from low to intermediate low band depths values.

R540: IH–H

SSBI: I–IH

BDI: L–IL

BDII: L–IL

No evident correlation. To be discussed during the data integration.
#9 (Coral)

Mainly in Apollo region, few pixels in Von Karman/Leibnitz

From intermediate high to high reflectance values, from low to intermediate low spectral slope values,  intermediate band depths values.

R540: IH–H

SSBI: L–IL

BDI: I

BDII: I

Crater walls
#10 (Sienna)

Apollo, Von Karman/Leibnitz

 

From intermediate high to high reflectance values, and spectral slope values, from intermediate to intermediate high band depths values.

R540: I–IH

SSBI: IH–H

BDI: I–IH

BDII: I–IH

No evident correlation. To be discussed during the data integration.
#11 (Dark Yellow)Apollo, Von Karman/LeibnitzFrom intermediate high to high reflectance values, from low to intermediate low spectral slope values, intermediate band depths values.

R540: IH–H

SSBI: L–IL

BDI: I

BDII: I

No evident correlation. To be discussed during the data integration.
#12 (Light Purple)

Mainly in Apollo region, few pixels in Von Karman/Leibnitz

From intermediate high to high reflectance values, from low to intermediate low spectral slope and band depths values.

R540: IH–H

SSBI: L–IL

BDI: L–IL

BDII: L–IL

High reflectance cratered terrains
#13 (Light green)Von Karman/LeibnitzFrom intermediate low to intermediate reflectance values, from intermediate high to high spectral slope values, from intermediate low to low 1 µm band depths values, from low to intermediate low 2 µm band depths values.

R540: IL–I

SSBI: IH–H

BDI: IL–I

BDII: L–IL

No evident correlation. To be discussed during the data integration.
#14 (Black)

Unclassified  Pixels Unit

All the regions that are not highlighted by the set of spectral parameters and so not included in the above units. Mainly associated with noisy data, indicative of residuals of calibration and instrumental artefacts.-----


Case 2: Apollo basin and Von Karman/Leibnitz craters combined analysis

Here we derived the spectral units of Apollo basins and Von Karman/Leibnitz regions while considering these two areas as a single region.

Figure 25:  Spectral parameters histograms of Apollo basins and Von Karman/Leibnitz regions. Blue lines indicate the mode values, while the green lines represent the class threshold values. BDI histogram in panel b) shows a double distribution value. In this case we did not consider the mode as a reference value but the value of separation of the two distributions.


Figure 26: SSBI vs R540 density scatter plot, compared with their distribution values. Colours of the threshold lines are consistent with the plots shown in the previous section.

Figure 27: BDI vs BDI density scatter plot, compared with their distribution values. Colours of the threshold lines are consistent with the plots shown in the previous section.



                          

Parameter

R540

Red (L)

Green (IL)

Blue (I)

Cyan (IH)

Magenta (H)

Mode = 0.0444

< 0.03250.0325–0.04000.0400–0.05200.0520–0.0620> 0.0620

#pixels

(2257374)

92077

(4.1 %)

343613

(15.2 %)

1126658

(49.9 %)

499621

(22.1 %)

195405

(8.7 %)


Figure 28:  Panel a) R540 class maps of Apollo basin (top) and Von Karman/Leibnitz craters (bottom). In panels b), c), and d) we show for each class, the average reflectance spectra, the normalised spectra at 540 nm and the continuum-removed spectra respectively.



Parameter

SSBI

Red (L)

Green (IL)

Blue (I)

Cyan (IH)

Magenta (H)

Mode = 1.51 1/µm

< 1.34 1/µm1.34–1.43 1/µm1.43–1.64 1/µm1.64–1.73 1/µm> 1.73 1/µm

#pixels

(2256209)

287321

(12.7 %)

389713

(17.3 %)

1362067

(60.4 %)

187861

(8.3 %)

29247

(1.3 %)


Figure 29:  Panel a) SSBI class maps of Apollo basin (top) and Von Karman/Leibnitz craters (bottom). In panels b), c), and d) we show for each class, the average reflectance spectra, the normalised spectra at 540 nm and the continuum-removed spectra respectively.


Parameter

BDI

Red (L)

Green (IL)

Blue (I)

Cyan (IH)

Magenta (H)

Mode = 0.054

< 0.0340.034–0.0620.062–0.0860.086–0.11> 0.11

#pixels

(2255384)

109017

(4.8 %)

741222

(32.9 %)

720858

(32.0 %)

483175

(21.4 %)

201112

(8.9 %)


Figure 30:  Panel a) BDI class maps of Apollo basin (top) and Von Karman/Leibnitz craters (bottom). In panels b), c), and d) we show the average reflectance spectra, the normalised spectra at 540 nm and the continuum-removed spectra for each class respectively.


Parameter

BDII

Red (L)

Green (IL)

Blue (I)

Cyan (IH)

Magenta (H)

Mode = 0.061

<0.0400.040–0.0490.049–0.0710.071–0.089> 0.089

#pixels

(2260486)

34807

(1.5 %)

183201

(8.1 %)

1177120

(52.1 %)

639961

(28.3 %)

225397

(10.0 %)

Figure 31:  Panel a) BDII class maps of Apollo basin (top) and Von Karman/Leibnitz craters (bottom). In panels b), c), and d) we show the average reflectance spectra, the normalised spectra at 540 nm and the continuum-removed spectra for each class respectively.


Figure 32:  Panel a) Apollo basin (top) and Von Karman/Leibnitz craters (bottom) SUs. In panels b), c), and d) we show the average reflectance spectra, the normalised spectra at 540 nm, and the continuum-removed spectra for each class respectively. Black areas represent the unclassified units.


Apollo basin and Von Karman/Leibnitz spectral units definition: Case 2


Spectral UnitLocationSpectral IndicationSpectral Parameters RangeAssociated Morphostratigraphic Unit
#1 (Red)

Mainly in Apollo region, few pixels in Von Karman/Leibnitz

 

Low reflectance values, from low to intermediate low spectral slope values, from intermediate high to high band depths values.

R540: L

SSBI: L–IL

BDI: IH–H

BDII: IH–H

Dark smooth plains 
#2 (Green)

Apollo, Von Karman/Leibnitz

 

Intermediate low reflectance values, from low to intermediate low spectral slope values, from intermediate high to high band depths values.

R540: IL

SSBI: L–IL

BDI: IH–H

BDII: IH–H

Intermediate dark smooth plains
#3 (Cyan)Apollo, Von Karman/LeibnitzFrom intermediate low to intermediate reflectance values, from intermediate to intermediate high spectral slope values and from  intermediate high  to high band depths values.

R540: IL–I

SSBI: I–IH

BDI: IH–H

BDII: IH–H 

No evident correlation. To be discussed during the data integration.
#4 (Blue)Apollo, Von Karman/LeibnitzFrom intermediate low to intermediate reflectance, spectral slope and band depths values.

R540: IL–I

SSBI: IL–I

BDI: IL–I

BDII: IL–I

Intermediate cratered terrains
#5 (Magenta)Apollo, Von Karman/LeibnitzFrom intermediate high to high reflectance values, from low to intermediate low spectral slope values, from intermediate high to high band depths values.

R540: IH–H

SSBI: L–IL

BDI: IH–H

BDII: IH–H

Bright ejecta
#6 (Maroon)

Mainly in Apollo region, few pixels in Von Karman/Leibnitz

From intermediate high to high reflectance values, from intermediate to intermediate high spectral slope values and band depths values.

R540: IH–H

SSBI: I–IH

BDI: I–IH

BDII: I–IH

No evident correlation. To be discussed during the data integration.
#7 (Sea Green)Apollo, Von Karman/LeibnitzIntermediate reflectance and spectral slope values, from intermediate high to high values and band depths values.

R540: I

SSBI: I

BDI: IH–H

BDII: IH–H

Intermediate reflectance ejecta and terrains 
#8 (Purple) 

Mainly in Apollo region, few pixels in Von Karman/Leibnitz

From intermediate high to high reflectance values, from intermediate to intermediate high spectral slope values, from low to intermediate low band depths values.

R540: IH–H

SSBI: I–IH

BDI: L–IL

BDII: L–IL

No evident correlation. To be discussed during the data integration.
#9 (Coral)

Apollo, Von Karman/Leibnitz

From intermediate high to high reflectance values, from low to intermediate low spectral slope values, intermediate band depths values.

R540: IH–H

SSBI: L–IL

BDI: I

BDII: I

Crater walls
#10 (Sienna)

Mainly in Apollo region, few pixels in Von Karman/Leibnitz

From intermediate high to high reflectance values, and spectral slope values, from intermediate to intermediate high band depths values.

R540: I–IH

SSBI: IH–H

BDI: I–IH

BDII: I–IH

No evident correlation. To be discussed during the data integration.
#11 (Dark Yellow)Apollo, Von Karman/LeibnitzFrom intermediate high to high reflectance values, from low to intermediate low spectral slope values, intermediate band depths values.

R540: IH–H

SSBI: L–IL

BDI: I

BDII: I

No evident correlation. To be discussed during the data integration.
#12 (Light Purple)

Mainly in Apollo region, few pixels in Von Karman/Leibnitz

From intermediate high to high reflectance values, from low to intermediate low spectral slope and band depths values.

R540: IH–H

SSBI: L–IL

BDI: L–IL

BDII: L–IL

High reflectance cratered terrains
#13 (Light green)Von Karman/LeibnitzFrom intermediate low to intermediate reflectance values, from intermediate high to high spectral slope values, from intermediate low to low 1 µm band depths values, from low to intermediate low 2 µm band depths values.

R540: IL–I

SSBI: IH–H

BDI: IL–I

BDII: L–IL

No evident correlation. To be discussed during the data integration.
#14 (Black)

Unclassified  Pixels Unit

All the regions that are not highlighted by the set of spectral parameters and so not included in the above units. Mainly associated with noisy data, indicative of residuals of calibration and instrumental artefacts.

------


Mars


The approach applied to Mars target areas is different from that considered for Mercury and the Moon. Mars' CRISM dataset is characterised by a higher spatial resolution with only partial coverage of the target sites preventing a global study but allowing a detailed local analysis of, for example, Crommelin crater and Arsinoes Chaos. Although the spectral parameters and detection limits or colour variation are reproduced considering the literature (Viviano-Beck et al. 2014, and http://crism.jhuapl.edu/data/mica/), due to the variability of Martian geology and CRISM acquisition conditions over time, we cannot perform a global scale comparison with our products. Here we follow the approach indicated by MICA file, which allows for a comparison case to case within different CRISM cubes, even though we are not able to define absolute value ranges from different images, for the case under consideration.

We examine two sites that are morphologically, stratigraphically and structurally analysed by other working groups.

1) Crommelin is a 114-km diameter crater within Arabia Terra with a large central bulge and well-preserved stratification with landforms and structures that can be interpreted as fold sets (Pesce et al. 2018). Six images cover different portions of the crater, from the central bulge to its floor. The spectral information indicates the non-mafic mineral phases in the crater, but diagnostic absorption features related to hydrated phases are not observed (not related to deposition or to hydrothermal      activity). Nevertheless, the dataset shows some colour variations in a specific RGB combination (FEM, although as described in the D4.2 deliverable we change the colour combination with respect to the one proposed by Viviano-Beck et al 2014) likely associated with different features within the crater.

2) Arsinoes Chaos is a region located east of Valles Marineris and west of Arabia Terra. For this area, we studied three images; one within the limit of the Top Unit and the Light-toned Layered Unit, and two images located to the northeast within one of the basement units of this crater. The first image highlights the differences between the two units with a CHL colour combination (Viviano-Beck et al. 2014, and http://crism.jhuapl.edu/data/mica/), indicating that the Light-toned Layered Unit could be characterised by hydrated material, similar to what has already been identified in other chaos terrains (e.g. Aram and Aureum Chaos, see also Luzzi et al. JGR Planets under revision and 2020 pre-print at https://eartharxiv.org/td297/). Unfortunately, ratioed spectra, which are useful for investigating the present mineral phases more thoroughly, cannot be extracted from these data. The other two images can be used to determine the bedrock composition of this chaos terrain, indicating possible phases associated with mafic mineral alteration, already observed in one small region. From these images we cannot exclude the possibility that parts of this unit elsewhere could include analogous altered material. It is possible to estimate the bedrock composition thanks to the threshold of two specific mineral parameters (LCPINDEX2 for pyroxene on the basaltic bedrock, and D2300 for the Band Depth at 2.3 um) and extrapolate their ratioed spectra (see Luzzi et al. JGR Planets under revision and pre-print at https://eartharxiv.org/td297/).

Mars case 1: Crommelin crater


imagesR channelG channelB channel
frs00029e820.0068–0.01740.1242–0.19510.1397–0.1929
frs00029fb50.0044–0.01050.1584–0.22590.1570–0.2117
hrs00012a180.0051–0.01030.2256–0.27140.1881–0.2561
frt0001fd590.0071–0.01400.1533–0.21510.1628–0.2188
frt00017c940.0080–0.02270.1287–0.20940.1229–0.1950
hrl000064b2  0.0065–0.0189 0.1244–0.1716 0.1696–0.2351

The values of limit of the RGB channel for the considered images are calculated at 2% limits from the histogram for each channel. 


Spectral UnitProposed nameSpectral CharacteristicsSpectral Parameters RangeAssociated Morphostratigraphic Unit
#1mafic glassy sand

They show no mafic absorption, nor the presence of hydrated phases nor carbonates.

The visible spectrum shows that considering the FEM colour combination (Viviano-Beck et al. 2014) this terrain is generally pink to red once we considered a distribution within the 2% of the RGB histogram values.

With the proposed colour combination this unit could indicate the presence of mafic minerals (less oxidised material) such as those found in volcanic sand (often glass–bearing, explaining the absence of pyroxene or olivine absorptions).

FEM colour combination, modified in the order with respect to Viviano-Beck et al 2014.

R: BDI1000VIS
G: BD530_2
B: SH600_2

The limits vary from image to image due to the different conditions of acquisition (geometry, martian day, noise, etc, see values in above table).



Low topography, with sand exposure.
#2ferric exposed dusty rock material

They show no mafic absorption, nor the presence of hydrated phases nor carbonates.

The visible range of the spectra permits to highlight that, considering the FEM color combination (Viviano-Beck et al. 2014) and a distribution within the 2% of the RGB hystogram values, this terrain appears generally green to blue.

With the proposed colour combination, green to blue indicates the probable presence of ferric material (maybe dust covering the exposed rocks). This can be affected by textural effects.

FEM colour combination, modified in the order with respect to Viviano-Beck et al 2014.

R: BDI1000VIS
G: BD530_2
B: SH600_2

The limits vary from image to image due to the different conditions of acquisition (geometry, martian day, noise, etc, see values in above table).

High topography. Exposed rocks are present.
#3Intermediate regionShows colour variation between the other classifications and is only a transitional portion, affected by superposition of information in the same pixel and/or photometric residual.

Not associated with geological units but it represents a transition between geological units, due to the lower resolution of hyperspectral images.

Contact blurred by the lower resolution of hyperspectral images and photometric residuals since no DTM was used to correct these data.

Mars Case 2: Arsinoes Chaos

Spectral UnitProposed nameSpectral IndicationSpectral Parameters RangeAssociated Morphostratigraphic Unit
#1Cap UnitPartially brighter region with respect to the other portion of the image.

RGB "CHL" white colour, brighter region with no indication of possible mineralogy but separate from the stratigraphically lower unit (#2)

Plateau on top of this Chaos, defined as Cap Unit.

CRISM cube frt00008233

#2Hydrated mineralogyIndication of a possible hydrated mineralogy but no clear absorptions in ratioed spectra could be extracted due to detection limits and noise in the image. No mafic mineralogy appears to be present.

RGB "CHL" yellow colour which can be indicative of hydrated mineralogy

Can be associated with the presence of the Light-toned Layered Unit, analogous with other chaos terrains on Mars closer regions. In this other chaotics minerals like detected mixtures of sulfates and phyllosilicates have being identified using the TES and THEMIS datasets or CRISM (Glotch & Christensen 2005; and Lichtenberg et al. 2010) .

CRISM cube frt00008233.

#3Bedrock material

Presence of mafic signature, with large 1 and 2 µm absorptions. Ratioed spectral signatures show an asymmetry in the  1µm band that could suggest the presence of a second mafic minerals like olivine or high-Ca pyroxene. 

Px Band Depth: LCPINDEX2 >0.02 



Mafic evidence which can be associated with possible low-Ca Pyroxene.

This evidence is within the ChF (chaotic terrains - fractured plains) unit even if the detection images are between the border of Arsinoes and Aureum Chaos.

CRISM cube frt000196b0_07 and frt00023790_07.

#4Bedrock alterationPresence of hydrated signatures at 1.9 and 2.3 µm.

Band depth at 2.3 µm (D2300) > 1% up to 4%

Evidence of smectite, as an alteration phase within some bedrock material (ChF unit) with light cyan colour in the PFM RGB.

CRISM cube frt000196b0_07.

References


Bernknopf, R. L., et al., 1993, Societal Value of Geologic Maps, USGS Circular 1111.

Glotch, Timothy D, & Christensen, P. R. 2005. Geologic and mineralogic mapping of Aram Chaos: Evidence for a water-rich history. Journal of Geophysical Research: Planets, 110(E9).

Ivanov, M.A., et al., 2018, Geologic History of the Northern Portion of the South Pole‐Aitken Basin on the Moon. Journal of Geophysical Research: Planets, Volume 123, Issue 10. https://doi.org/10.1029/2018JE005590

Lichtenberg, K. A., et al. 2010. Stratigraphy of hydrated sulfates in the sedimentary deposits of Aram Chaos, Mars. Journal of Geophysical Research, 115(E6), E00D17. https://doi.org/10.1029/2009JE003353

Luzzi et al, under review. Tectono-magmatic, sedimentary and hydrothermal history of Arsinoes and Pyrrhae Chaos, Mars. Journal of Geophysical Research: Planets.

Luzzi et al. 2020. Tectono-magmatic, sedimentary and hydrothermal history of Arsinoes and Pyrrhae Chaos, Mars. https://eartharxiv.org/td297/ preprint doi: 10.31223/osf.io/td297

Pesce et al. 2018. Structural Mapping of the Inner Layered Deposits of the Crommelin Crater (Mars). EPSC2018-294. https://meetingorganizer.copernicus.org/EPSC2018/EPSC2018-294.pdf

Rothery et al., 2020, Icarus submitted.

Viviano-Beck, C. E.,  et al 2014. Revised CRISM spectral parameters and summary products based on the currently detected mineral diversity on Mars. Journal of Geophysical Research: Planets, 119(6), 1403–1431. 852

The Mica Files. John Hopkins University. http://crism.jhuapl.edu/data/mica/


  • No labels