Simulated Rover Traverse at Silver Lake Playa, California:

Field Trip Report

December 15, 1997

Participants
Objectives
Approach
Silver Lake Study Site Description
Camera Description
Overview of Stations
Detailed Description of Stations

Participants

Objectives

One of the long term challenges for future planetary robotic missions is to develop and use technology to enable autonomous analyses of rover data. The objective of this field trip was to collect field data and samples to simulate data sets acquired by an autonomous rover. A related objective is to acquire relevant laboratory measurements on the samples using a suite of instruments analogous to the instrument package on the field prototype of the Mars rover, FIDO (Field Integration Data and Operations). Silver Lake, California (Figure 1), was selected as the field location because it is a paleolake site with preserved beach ridges, recessional shorelines and a subaqueous depositional structure (tufa bar). Field-based images were acquired along a traverse and at four other stations at the site. The image and laboratory data sets will be used to develop and test pattern recognition algorithms designed to identify signatures indicative of lacustrine environments. In addition, the data in this study will be organized using the GIS software package Spans Explorer and will be posted on the internet in an interactive website. The results of this simulated rover exercise provide valuable lessons for mission operations. In addition, the data obtained in the laboratory will yield data and experience for using remote sensing in environments where rocks are obscured by surface coatings such as desert varnish.  

Approach

For the simulated rover traverse, a range of different type images were synthesized. Satellite (Landsat Thematic Mapper (TM); Figure 2) and airborne (Thermal Infrared Multispectral Scanner (TIMS); Figure 3) images of Silver Lake provide a regional geologic context for the site and are analogous to data sets obtained from an orbiter. To simulate the images obtained by a rover, stereo and monoscopic digital cameras were used to acquire field images. In addition to images which documented the large-scale features of the field site, close-up images of samples in-situ were obtained prior to collection. Position data for sites was supplied by a portable Magellan global position system (GPS). Samples were collected at most stations along the traverse and at the isolated field stations. Future work will include Mössbauer, Raman and infrared spectroscopy to complete the simulated rover transect data set. Petrographic analyses of the field samples will be conducted for calibration purposes.  

Silver Lake Study Site Description

Silver Lake playa is located north of Baker in the Mojave Desert of California (Figure 1). It supported lakes during Pleistocene glacial epochs. The setting is similar to the proposed landing sites for the 2001 rover mission. Preserved beach ridges, recessional shorelines, a subaqueous depositional structure (tufa bar) and a break-out channel are all morphological evidence of the paleolake. The playa surface is composed of silt-sized particles with predominately clay mineralogy. The playa is bounded to the west by the Soda Mountains, which are composed of a suite of varnished rock-types (Precambrian metasedimentary rocks, Lower Cambrian quartzite units, and Pre-Tertiary plutonics which range in composition from gabbro to granite) and uncoated rocks (Lower Permian and Lower Cambrian/Precambrian carbonates).  

Camera Descriptions

The digital camera used in this exercise to capture wide-angle and close-up images was a Kodak DC210 Zoom. The 24-bit color images have a picture resolution of 640 X 480 pixels and were saved in JPG file format. For stereo images, the Kodak DCS-400 system was used. The field of view for each camera is 30 degrees. Images acquired with the stereo DCS-400 cameras have a picture resolution of 962 X 506 pixels.  

Overview of Stations

The localization of carbonate at the tufa bar would easily be identified from orbital and descent imaging as a target site. In the Landsat TM false-color image of Silver Lake, with bands 2 (0.52 – 0.60 m m), 4 (0.76 – 0.90 m m), 7 (2.08 – 2.35 m m) mapped as blue, green and red respectively, carbonate units are pale green in color (Figure 2). Carbonate units are green in a TIMS false color image of Silver Lake with decorrelation stretched bands 1 (8.2 – 8.6 m m), 3 (9.0 – 9.4 m m), 5 (10.2 – 11.2 m m) mapped as blue, green and red (Figure 3). Based on these images, a traverse was planned from the playa southwest-ward to the tufa bar (stations 1-6). The traverse crossed a set of recessional shorelines via a series of six waypoints (Figure 4). At each station, images were acquired to document the traverse. Samples were collected from four of the six waypoints.

Upon completion of the simulated rover traverse, other stations at the site were visited (Figure 1). At station 7, examination of the playa surface revealed dessication mudcracks. At the break-out channel to the north of Silver Lake, one station (station 10) was selected to document the regional features of the channel while two other stations (stations 8 and 9) documented the differences in rock exposures on either side of the channel. On the eastern bank of the channel is a carbonate-cemented conglomerate while to the west is a gneissic outcrop with tufa cap. Samples from both outcrops at these sites were collected.

Finally, a side-trip was taken to the Granite Mountains (station 11) to document a weathered granite surface. The purpose of this station was to introduce the JPL participants to variations in surface texture, which may be an important factor in pattern recognition algorithms. No sample was obtained at this site.  

Detailed Description of Stations

Note: Table 1 contains a detailed catalog of location, instrument orientation and data acquisition. Table 2 describes the seven field samples.  

Simulated Rover Traverse

Station 1: Approach to Tufa Bar

From this vantage point, the tufa bar target (station 6) is visible in the background (Figure 5; DC210 image 117). The traverse direction to the tufa bar is approximately S75° W via six waystations. In the midground a concentration of rocks was selected as station 2, the rock garden. The playa surface is in the foreground. Shorelines are not discernable in this image.

Station 2: Rock Garden

At station 2, station 3 (regional shorelines) and station 6 (tufa bar) are visible. Faint shorelines are also visible, but are not laterally continuous (Figure 6; DC210 image 120). The tufa bar acts as a barrier to erosion and protects these fragile geomorphic features. Desert pavement (a tight mosaic of cobbles) surrounds a variety of angular boulders dubbed the "Rock Garden". All rocks appear to have desert varnish, a dark surface coating that smoothes the exterior of the rock. Figure 7 (DC210 image 119) is a close-up image of the Rock Garden. A sample was collected from the prominent center rock in Figure 7. A macro-image of the post-sampled rock is shown in Figure 8 (DC210 image 124) where the light-colored, fresh powder is on top of the varnished exterior surface of the rock.

Station 3: Regional Shorelines

In Figure 9 (DC210 image 121), three laterally continuous, distinct shorelines are discernable with fainter shorelines evident in the background. Station 3 (located atop the second shoreline) and station 6 (tufa bar) are identifiable in this image. A desert pavement comprised of varnished cobbles covers the ground.

Station 4: Shoreline Bench #2

Station 4 is atop the second prominent shoreline (foreground) identified at station 3 (Figure 10; DC210 image 128). The third shoreline is easily discernable. Again, the tufa bar (station 6) is located in the background. Four viewing angles were acquired to illustrate the effects of photometry on image quality (Figure 11). (Note the bone-shaped red rock in the center, which is present in each image in this exercise.) Figure 12 (DC210 image 122) illustrates the site viewed to the west. Figure 13 (DC210 image 123) was acquired from a position 45° northward of Figure 12 and is looking southwest. Continuing in a counterclockwise fashion, the instruments were moved another 90° to view the site to the southeast (Figure 14; DC210 image 125). This illustrates the effects of forward scattering (specular reflection) which provides textural information about rock surfaces. The mirror like reflections from rocks suggests the presence of surface coatings. Figure 15 (DC210 image 126) is looking approximtely due north and has an orientation of almost zero phase angle. The back scattered light produces a very bright image that looks "washed-out". Figure 16 (DC210 image 127) is a macro-image of an etched carbonate cobble collected from station 4.

Station 5: Varnished Rocks and Ventifact

Figure 17 (DC210 image 132) looks toward the wave-cut tufa bar (station 6). At station 5, an array of varnished rock surfaces ranging from pebbles to angular boulders formed a desert pavement (Figure 18; DC210 image 129). A macroscopic view of the varnished ventifact (a wind-polished rock) was acquired (Figure 19; DC210 image 133); a portion of this gabbro sample was collected. Again, photometric effects were documented (Figure 20): specular direction (Figure 21; DC210 image 130) and back scattered direction (Figure 22; DC210 image 131).

Station 6: Tufa Bar

At the tufa bar, angular plutonic rock fragments are bound together by calcium carbonate cement (Figure 23; DC210 image 134). Wave action produced the notch in the tufa bar. A sample of the tufa bar was collected. A view back along the traverse was also acquired (Figure 24; DC210 image 135). Note that the shorelines are not discernable when viewing northeast. The playa edge and distant Silurian Hills are visible in the background.

Station 7: Silver Lake Playa

The playa surface is comprised of silt-sized clay particles. Dessication of the playa has resulted in shrinking of the clay particles resulting the formation of mudcracks, as seen in the foreground of Figure 25 (DC210 image 136). In Figure 26 (DC210 image 137), a closer view reveals the mudcracks form a polygonal mosaic. At a macroscopic view (Figure 27; DC210 image 138), the fine-scale surface texture and crack initation points of mudcracks is visible.

Northern Stations

Station 8: Conglomerate on east channel bank

Regional view of conglomerate on east side of outflow channel (Figure 28; DC210 image 139) has large-scale cavities. At a closer-range, the finer scale features and smaller cavities are visible (Figure 29; DC210 image 140) At the outcrop, ranges in cobble size and the interstitial carbonate cement are observed (Figure 30; DC210 image 141). Conglomerate fragments and cement were collected.

Station 9: Gneiss and tufa on west channel bank

On the western bank of the outflow channel is a gneissic outcrop with a tufa cap (Figure 31 and Figure 32; DC210 image 143, 142). Note the talus block is in an ideal position for a rover to examine (Figure 33; DC210 image 144). Macroscopic imaging of the talus block reveals the carbonate cement (Figure 34; DC210 image 145). A portion of the talus block was collected.

Station 10: Regional Channel View

Figure 35 (DC210 image 146) is an image acquired from the channel bed viewing to the north. Sand bars are aeolian (wind-blown) deposits.

Granite Mountains

Station 11: Granite Outcrop

The team visited the Granite Mountains as an auxillary stop to illustrate the textural properties of weathered granite. Granite weathers by granular disintegration whereby the rock is broken down grain-by-grain. Figure 36 (DC210 image 147) illustrates a granite boulder. The macroscopic image (Figure 37; DC210 image 148) shows the individual mineral crystals of the weathered granite surface.