Detecting and mapping ancient soils

During the early Cenozoic, and perhaps before that, huge areas of the exposed continental surface were subject to hot humid climatic conditions. That broke down every conceivable rock type to a few simple minerals that were both stable and insoluble. Such intense weathering possibly affected 30% of the land area during those ‘hothouse’ times. Where the surface was flat, the resulting residual soils were preserved to form laterites, strongly layered mineralogically. Since one of the common components is bright-red hematite, and its brown hydrous equivalent goethite, and another is brilliant white kaolinite, laterites are also stunningly layered in colour from white iron-poor clays at their base through an middle mottled yellow, orange, pink and white zone, to brick-red iron-rich ferricrete at the top of the sequence. No-one can fail to see laterites where they are exposed, but few geologists have set out to understand them. A recent paper provides a clear guide to begin that work on a grand scale, and also to chart where their unique properties and socio-economic pros and cons can be developed or avoided respectively (Andrew Deller, M.E. 2006. Facies discrimination in laterites using Landsat Thematic Mapper, ASTER and ALI data—examples from Eritrea and Arabia. International Journal of Remote Sensing, v. 27, p. 2389–2409).

The key to the long and complex chemical and mineralogical evolution of laterites lies in the different layers or facies in these palaeosols. Because they are thin and once present over vast areas of Africa, South America, India and Australia, their presence or absence today is a guide to the history of erosion and intraplate deformation after they formed. Each facies has very different chemical and physical properties, some advantageous, and some decidedly a threat of some kind, recognised and well documented by M.E. Andrews Deller of the British Open University. For instance, the clay zone is a lubricant that can encourage landslides of great thicknesses of overlying rock, yet is a potential resource — it is China Clay. Hard and porous ferricrete, containing both iron minerals and clays, makes it a cheap source of bricks and even road aggregate. But hematite can pose a frightening risk. Its open structure soaks up dissolved ions, including infamously those of arsenic, which lateritisation can set in motion from the rocks on which it develops. Hematite dissolves under reducing conditions, and should those develop on old laterites arsenic might be liberated to groundwater. Another associated compound that laterites can release is magnesium sulfate (Epsom Salts), an natural emetic but also a potential remedy for eclampsia that threatens mothers and their babies throughout laterite-mantled Africa.

Andrews Deller’s paper is a mine of laterite-related information, yet its central theme is the essential first step of mapping them and discriminating their facies. Her starting point is their mineralogical simplicity, and the unique and distinct spectral properties of those constituent minerals. She matches these to the spectral coverage of freely available remote sensing data — Landsat TM, ASTER and ALI — each of which offers nuances to be exploited in uniquely discriminating the zones. Rather than setting out to ‘unveil’ sophisticated new methods of computer analysis (to which few in laterite-encrusted areas would have access), she chose the simplest useful approaches to a previously overlooked challenge: laterite facies have never been discriminated and mapped before. The results in this well-illustrated paper are stunning, and any geologist, and quite likely many lay people can understand what they show, thanks to careful discussion. The result is a paper that combines interest, novelty and usefulness. The last is the best aspect: geologists can learn from the paper how confidently to make highly informative maps cheaply and quickly.

ASTER data and earthquakes

NASA’s Jet Propulsion Laboratory in Pasadena, California is a huge engine of across-the-board innovation. In my field, remotely sensed geology, everyone pounces eagerly on publications by its scientists because they are bound to push techniques and applications forwards, often in surprising contexts, such as archaeology from space. One such nugget is about to be published (probably this month) in the premier geoscience journal EPSL (Avouac, P. et al. 2006. The 2005, Mw 7.6 Kashmir earthquake: Sub-pixel correlation of ASTER images and seismic waveforms analysis. Earth and Planetary Science Letters, in press doi:10.106/j.epsl.2006.06.025) and amply justifies my impatient preview here. It offers great potential for monitoring the effects of natural hazards that involve mass motion using free (for bona fide researchers and, hopefully, humanitarian organizations) satellite image data.

Jean-Phillipe Avouac and colleagues at JPL applied a well-tried approach in remote sensing — comparison of images captured on different dates—in trying to assess the extent and magnitude of ground motion involved in the 8 October 2005 Kashmir earthquake that claimed at least 80 thousand lives. But theirs is a before-and-after study with a revolutionary new slant. ASTER data from the joint US-Japanese Terra satellite resolves the ground with a resolution as sharp as 15 m, in several wavebands of EM radiation. In their own right, these bands contain huge amounts of information about vegetation, rocks and soils, and many other environmental attributes. Particularly with vegetation, comparing data from different years or seasons soon shows up changes and clues as to why they occurred. But ASTER has another potential view to offer. Two of its sensors, one pointing vertically downwards, the other obliquely back along its ground track, constitute a stereopair. They can be viewed together to give dramatic 3-D visualizations of terrain. With the appropriate software, the parallax difference between the location of each point on the ground in the two images produces a map of terrain elevation. The novelty and potential in Avouac et al. is to combine ASTER data from two instants in time to find places that have shifted in position in the meantime. So that they match geographically, they used stereo-derived terrain elevation to remove geometric distortions caused by viewing rugged relief with effectively a wide-angle camera. The key to extracting deformation parameters is applying shape-detection software to images from before and after an event, and then finding the magnitude and direction of the differences between landform shapes to chart movement. The 15 m resolution poses a limit, but the sophistication of the algorithms enables shifts of the order of less than a metre to be detected at a coarse resolution of 150 m. But that is quite sufficient to show what happened in Kashmir along the entire length of fault movement in 2005. Applied to commercially available stereo data (up to 0.65 m resolution) the results would be awesome.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s