Registering Planetary Datasets for Data Fusion: A “Force Multiplier” for Planetary Science

B. A. Archinal, K. L. Edmundson, R. L. Kirk, and L. R. Gaddis

U. S. Geological Survey, 2255 N. Gemini Drive, Flagstaff, AZ 86001, USA, barchinal@usgs.gov

47th Lunar and Planetary Science Conference

The Woodlands, Texas

2016 March 23

Introduction: A “Force Multiplier”

Putting pixels in the right place…for Science and Exploration!

o Planetary data are only useful when they are converted into knowledge

o Registration in a common coordinate system at known levels of accuracy is essential for:

o Study of broad areas (bigger than a single observation), including distribution

o “Data fusion”: comparison and combination of multiple data sets

o Enhanced by proper photometric processing and projection onto topography

o Needed & useful for all remotely -sensed data, singly or in combinations:

o monochrome, color, multispectral, hyperspectral, radar, or lidar

o Multiplies the value of registered data by orders of magnitude, promoting

more understanding and more science per dollar

RS Data

Metadata

Prior Data as control

Geodetic

Control

Orthorect-

ification

Calibration

Precision RegisteredProducts

Topography

Science analysis

KNOWLEDGE

Other prec. products

Precision Cartography Has Many Parts

o Geodetic control—The biggest single piece

o Improving consistency and accuracy of spacecraft trajectory and pointing, etc.

o Quantifying the uncertainty

o Other pieces of the puzzle

o Calibration (detailed knowledge of instrument geometry)

o Topography (knowledge of the target geometry) to remove parallax distortions

o Specification and use of standard coordinate reference system and frame

o These parts of the process all interact

o Calibration can be refined as part of the geodetic control process (“self-calibration”)

o Data used to generate topographic models must also be controlled

o Coordinate frames are defined via creation of an initial control solution

Registering data – Geodetic Controlo Usually provides significant improvement over dead reckoning

o i.e., positioning data based on a priori geometry data

o Consists of performing least squares photogrammetric, radargrammetric, and/or altimetric solutions to register data into a common frame with known levels of accuracy

o The need for control solutions is critical for all missions

o Early missions with low accuracy a priori data

o Current and future missions with massive, high-spatial resolution data sets requiring substantially better instrument and target geometry information for registration

o Only control solutions support proper assessment of the relative and absolute uncertainty in the position of the data

o A critically important but often overlooked element of data fusion that helps to optimize its use for planetary exploration

“Seeing is believing – measuring is knowing”

(with uncertainties)- Norman Henderson, Henderson Aerial Surveys

Critically Needed ToolsMany mapping tools need improved or created to process existing and planned future datasets.

Examples:a) better tie-pointing methods and statistical outlier detection algorithms

b) algorithms and software to efficiently process and store massive data sets

c) new methods to jointly tie together image, radar, and altimetry data

d) photoclinometric methods with error analysis capabilities

e) new mapping methods for small and irregular bodies

f) speed improvements (real-time processing for operations and mission science)

There is value and economy of scale in having all data controlled in a single effort, both to provide the benefits listed here and to allow processing to be done most efficiently by experts using the most rigorous algorithms and software.

6 images to be controlledImagine you had tens of

thousands! [From K. Edmundson, after

Kraus (1993) “Photogrammetry…”.]

Let’s control IR (left) to RADAR! (right)[From R. Kirk]

High-Value Examples Highlighting Precision Cartography - 1

o Surface topography of the lunar Apollo 17 landing site to properly project and photometrically correct Clementine UV images to substantially improve our understanding of the minerals, stratigraphy, and context of samples collected from that site

Robinson and Jolliff (2002) JGR 107, E11, 5110

For figures, see Figures 4 and 8 in Robinson

and Jolliff (2002) JGR107, E11, 5110.

High-Value Examples Highlighting Precision Cartography - 2

Use of radargrammetry solutions and topography to control LRO Mini-RF images of the lunar poles to support accurate identification and mapping of volatiles such as water

o Rectification of radar image of Shackleton crater Mini-RF on LOLA (left/before; right/after) =>

o Comparison of uncontrolled (left) and controlled (right, with east-looking mosaic shown in cyan, west-looking in red) Mini-RF mosaics for Hermite A, a 20-km crater. Mismatches within and between mosaics are ~1-3 km uncontrolled, <30 m after control adjustment.

Kirk et al. (2013) LPS XLIV #2920.

High-Value Examples Highlighting Precision Cartography - 3

Recurring Slope Lineae: Dynamics, distribution, composition—all from mapping

Ojha et al. (2015) Nature Geoscience, 8, 829

For figure, see Figure 1, panels a-c, in Ojha et al. (2015) Nature Geoscience, 8, 829

Image: NASA PIA14472

High-Value Examples Highlighting Precision Cartography - 4

o Remote and in-situ exploration of lunar polar volatiles, including use of multiple registered data sets to identify and map volatiles and other resources to support long-duration human exploration and eventually real-time science operations. Examples are depth to ice and days of sunlight

Heldmann et al. (2015) Advances in Space Res. 55, 2427

For figures, see Figures 1 and 2 in Heldmann

et al. (2015) Advances in Space Res. 55, 2427

High-Value Examples Highlighting Precision Cartography - 5

o Use of topography to model "clutter" in sounding radar data (MARSIS and SHARAD at Mars, RIME and REASON under development for the Jovian system, and small body radar tomography by future missions)

SPLD thickness, based on MARSIS and MOLA.

Orosei, et al. (2015) Planet. Space Sci. 112, 98

SHARAD observation of InSight landing site compared to clutter model — Borrowed from Than Putzig, SWRI

For figure, see Figure 8 in Orosei, et al. (2015)

Planet. Space Sci. 112, 98

High-Value Examples Highlighting Precision Cartography - 6

o Use of global control across multiple data sets over time to characterize the precise rotation state of the Enceladus and to infer the presence of a global ocean

o Control network points and physical libration amplitude

Thomas, et al. (2016) Icarus, 264, 37

Image: NASA PIA19058

For figure and Table, see Figure 1c and

Table 1 in Thomas, et al. (2016) Icarus, 264,

37

High-Value Examples Highlighting Precision Cartography - 7

o Blink between the average of 3 Titan ISS images with the controlled averaged mosaic, to show noise reduction and improvement in detail

o Blink of controlled ISS mosaic to RADAR SAR, showing visible/radar registration

Archinal, et al. (2013) LPS XLIV, #2957

Summary

o Planetary data must be registered in a common coordinate systemo Supports the high level of accuracy and effectiveness

enabled by the multiple high-quality data sets

o Allows derivation of the precision and accuracy of the location of the data

o Whether existing data or for future planetary science and exploration

o The precision cartography process outlined here must be a key element of any mission planning and execution

o Such processing holds the promise of substantially multiplying the value of planetary data and missions