Producing MIDS NAC DEMs from MESSENGER Images · Producing MESSENGER DEMs from MDIS NAC Images...

24
Producing MESSENGER DEMs from MDIS NAC Images MANHEIM 1 , HENRIKSEN, ROBINSON, AND THE MESSENGER TEAM 1 ARIZONA STATE UNIVERSITY, TEMPE AZ—[email protected]

Transcript of Producing MIDS NAC DEMs from MESSENGER Images · Producing MESSENGER DEMs from MDIS NAC Images...

Producing MESSENGER DEMs

from MDIS NAC ImagesMANHEIM1, HENRIKSEN, ROBINSON, AND THE MESSENGER TEAM

1ARIZONA STATE UNIVERSITY, TEMPE AZ—[email protected]

MESSENGER

Mercury Dual Imaging System (MDIS)

2 framing cameras: a monochrome NAC and a multispectral WAC NAC wasn’t a stereo camera, but off-nadir observations enable the

creation of DEMs

NACs have 5 m pixel scale at closest approach

Mercury Laser Altimeter (MLA)

Radial accuracy of < 20 m

Only available between 90° N and 18° S

Highly Elliptical Orbit Periapsis: 200 – 500 km (near North Pole)

Apoapsis: 10,000 – 15,000 km

Launched: August 2004

Mercury Orbit: March 2011

Completed: April 2015

Methodology: Overview

Site Selection & Image Selection

Illumination Conditions

Imaging Geometry

DEM Production

Using the USGS Integrated Software for

Imagers and Spectrometers (ISIS) and

SOCET SET 5.6

Error analysis

Creating data products for PDS release

Image Selection

Selecting stereo images

requires compromise

between finding optimal

images and building up

desired coverage.

These parameters will

dictate the precision of

the final product.

Strength of stereo: parallax between the two images / unit height

Illumination compatibility: distance between the tips of shadows in two images / unit height

*Guidelines adapted from Becker, et al. 2015.

Images Selected for Our DEMs n=49

Selecting Images to Build Mosaics

Sander Crater mosaic:

21 images

Selecting Images to Build Mosaics

Sander Crater mosaic:

21 images

36 stereo pairs (areas of overlap producing ‘good’ stereo)

Selecting Images to Build Mosaics

Sander Crater mosaic:

21 images

36 stereo pairs (areas of overlap producing ‘good’ stereo)

Resulting DEMs can be

mosaicked together

(using ISIS) to create one

large-area DEM

*This process is quite similar to that described for LROC NAC DEMs in Henriksen, et al. 2017.

Processing in SOCET SET: Overview

Import into SOCET SET 5.6

Relative Triangulation

Registration to MLA tracks

DEM Extraction

Orthophoto Generation

Creating Additional Data Products

PDS Release

Processing in SOCET SET: Overview

Import into SOCET SET 5.6

Relative Triangulation

Registration to MLA tracks

DEM Extraction

Orthophoto Generation

Creating Additional Data Products

PDS Release

*This process is quite similar to that described for LROC NAC DEMs in Henriksen, et al. 2017.

*This process is quite similar to that described for LROC NAC DEMs in Henriksen, et al. 2017.

Processing in SOCET SET: Overview

Import into SOCET SET 5.6

Relative Triangulation

Registration to MLA tracks

DEM Extraction

Orthophoto Generation

Creating Additional Data Products

PDS Release

Registration to MLA

Not always possible: for DEMs with little

coverage, or further south, there may be little or no MLA coverage.

In this case, find an overlapping stereo

pair of WAC images

Tie NACs to the WACs

WACs are manually registered to MLA

NACs and WACs are bundle-adjusted

together for absolute orientation

NAC

WAC

MLA

*This process is quite similar to that described for LROC NAC DEMs in Henriksen, et al. 2017.

Processing in SOCET SET: Overview

Import into SOCET SET 5.6

Relative Triangulation

Registration to MLA tracks

DEM Extraction

Orthophoto Generation

Creating Additional Data Products

PDS Release

Processing in SOCET SET: Overview

Import into SOCET SET 5.6

Relative Triangulation

Registration to MLA tracks

DEM Extraction

Orthophoto Generation

Creating Additional Data Products

PDS Release

*This process is quite similar to that described for LROC NAC DEMs in Henriksen, et al. 2017.

Processing in SOCET SET: Overview

Import into SOCET SET 5.6

Relative Triangulation

Registration to MLA tracks

DEM Extraction

Orthophoto Generation

Creating Additional Data Products

PDS Release

*This process is quite similar to that described for LROC NAC DEMs in Henriksen, et al. 2017.

Data Products

Sander DEM: 21 images, 36

stereo pairs

A readme file was also

produced for each DEM,

containing error analysis

results.

Error Analysis

Vertical precision (“relative linear error”)

is calculated by SOCET SET at a 90%

confidence level

expected to be < DEM pixel scale

DEMs Produced at ASU

Site Name

No. ofStereo Pairs

Center Latitude,

Longitude(°N, °E)

Mosaic Pixel

Scale (m)

RelativeLinear

Error (m)

MLA Mean Offset

(m)MLA Standard Deviation (m)

Catullus Crater* 1 21.88°, 292.5° 84 84 -255 188Cunningham Crater 9 30.40°, 157.0° 105 85 2 58Degas Crater 1 36.86°, 232.6° 97 70 -1 109

Hynek Scarp(unofficial name) 1 -31.16°, 82.5° 500 383 - -Kertesz Crater* 2 31.45°, 146.3° 120 98 -9 51Kuiper Crater 1 -11.3°, 329.1° 270 161 - -

Paramour Rupes* 3 -5.07°, 145.1° 450 261 -5 185

Raditladi Hollows 3 15.20°, 120.2° 180 137 -44 52Sander Crater 36 42.50°, 154.7° 162 70 -44 65

*Error analysis was performed on WAC images which were tied to the NAC DEMs.

Error Analysis

Positional Accuracy: Compare to MLA

Only possible for DEMs with overlapping

MLA tracks (northern hemisphere)

Ideally, offsets are < 20 m (the radial

accuracy of MLA)

DEMs Produced at ASU

Site Name

No. ofStereo Pairs

Center Latitude,

Longitude(°N, °E)

Mosaic Pixel

Scale (m)

RelativeLinear

Error (m)

MLA Mean Offset

(m)MLA Standard Deviation (m)

Catullus Crater* 1 21.88°, 292.5° 84 84 -255 188Cunningham Crater 9 30.40°, 157.0° 105 85 2 58Degas Crater 1 36.86°, 232.6° 97 70 -1 109

Hynek Scarp(unofficial name) 1 -31.16°, 82.5° 500 383 - -Kertesz Crater* 2 31.45°, 146.3° 120 98 -9 51Kuiper Crater 1 -11.3°, 329.1° 270 161 - -

Paramour Rupes* 3 -5.07°, 145.1° 450 261 -5 185

Raditladi Hollows 3 15.20°, 120.2° 180 137 -44 52Sander Crater 36 42.50°, 154.7° 162 70 -44 65

*Error analysis was performed on WAC DEMs which were tied to the NAC DEMs.

DEMs Produced at ASU

Site Name

No. ofStereo Pairs

Center Latitude,

Longitude(°N, °E)

Mosaic Pixel

Scale (m)

RelativeLinear

Error (m)

MLA Mean Offset

(m)MLA Standard Deviation (m)

Catullus Crater* 1 21.88°, 292.5° 84 84 -255 188Cunningham Crater 9 30.40°, 157.0° 105 85 2 58Degas Crater 1 36.86°, 232.6° 97 70 -1 109

Hynek Scarp(unofficial name) 1 -31.16°, 82.5° 500 383 - -Kertesz Crater* 2 31.45°, 146.3° 120 98 -9 51Kuiper Crater 1 -11.3°, 329.1° 270 161 - -

Paramour Rupes* 3 -5.07°, 145.1° 450 261 -5 185

Raditladi Hollows 3 15.20°, 120.2° 180 137 -44 52Sander Crater 36 42.50°, 154.7° 162 70 -44 65

*Error analysis was performed on WAC DEMs which were tied to the NAC DEMs.

PDS Release

All MDIS NAC DEMs and associated

products produced at ASU were

released to the public through the

PDS in December 2016.

PDS Release

All MDIS NAC DEMs and associated

products produced at ASU were

released to the public through the

PDS in December 2016.

https://pds-imaging.jpl.nasa.gov/volumes/mess.html

References[1] Becker, K. J. et al. (2015) LPS, 46, Abs. 2703.

[2] Henriksen, M. R. et al. (2017) Icarus, 283, 122-137.

[3] Cavanaugh, J. F. et al. (2007) Space Sci. Rev., 131, 451-497.

[4] Fassett, C. I. (2016) Planet. Space Sci., 134, 19-28.

[5] Zuber, M. T. et al. (2012) Science, 336, 217-220.

[6] Anderson, J. A. et al. (2004) LPS, 35, Abs. 2039.

[7] Becker, K. J. et al. (2016) LPS, 47, Abs. 2959.

[8] Forstner, W. et al. (2013) Manual of Photogrammetry (6th ed.), 785-955.

[9] BAE Systems (2011) SOCET SET User's Manual, v. 5.6.

[10] BAE Systems (2007) NGATE. White Paper.

[11] Zhang, B. (2006) ASPRS, p. 12.

[12] Miller, S. et al. (2013) Manual of Photogrammetry (6th ed.), 1009- 1043.

[13] Warmerdam, F. (2008) Open Source Approaches in Spatial Data Handling, 87-108.

[14] Hawkins, S. E., III et al. (2007) Space Sci. Rev., 131, 247-338.