Tuesday, October 07, 2008

Remote sensing of Afghanistan more complete than of U.S.

At this morning's State Geologist's mid-year meeting at the GSA conference in Houston, USGS director Mark Myers talked briefly about their work in Afghanistan. They completed a comprehensive hyperspectral remote sensing survey of the entire country, the largest such survey ever done in the world. Mark described how valuable this survey of Afghanistan is, for mineral exploration, for environmental issues, for a whole range of scientific studies. He then commented how wonderful it would be if we could get similar surveys in the states.

Isn't it ironic that we as a country can afford to undertake a national hyperspectral survey of Afghanistan because it's funded through national security funding channels but similar funding sources are not available to support domestic studies?

1 comment:

  1. Anonymous7:11 AM

    This is an interesting topic for discussion - I would even go so far as to say that I am fairly certain a comprehensive hyperspectral data set of a single US state is not publicly available. That is obviously very disappointing. Australia is probably the world leader when it comes to the total data coverage, whether it be traditional remote sensing-based, or of a geophysical nature. They are very good at collecting data and making it available for analysis by university or industry researchers.

    However, you must keep the context of landcover in mind when comparing Afghanistan or Australia versus the US. The bulk of Afghanistan is climatically classified as arid to semi-arid, and thus you do not have nearly the amount of vegetation covering the surface of the Earth (and the associated attenuation or scattering of the reflected or emitted energy of underlying minerals). A complete country-scale mineral survey using hyperspectral image data would therefore be massively inefficient. My point does not, however, change the fact that scientists interested in mapping or better understanding vegetation would obviously also enjoy having this resource.

    A final comment I will make about large-scale hyperspectral surveys is that it would be much easier to undertake if the customer were a single agency, etc. This is because the data can be treated so many different ways, and a particular scientist may have a preferred way to set collection, calibration, processing and analysis parameters. Obviously if you are collecting massive amounts of data for public consumption, you have to try to satisfy many different needs, which may turn out to be more trouble than it is worth. Worst-case scenario is that you collect the data and then it is invalidated because a particular scientist's parameters were not followed. Thus a single agency, in this case the USGS, can define its own parameters to follow, which will have been developed after conducting their own research to find out what works and what does not (and they are the experts!).

    I found more information on the Afghanistan data collect & analysis here: