Close Window

Remote sensing of terra preta in Amazonia

Michael Palace, CSRC-University of New Hampshire, michael.palace@unh.edu (Presenter)
Mark Bush, Biological Sciences, Florida Institute of Technology, Melbourne, FL, mbush@fit.edu
Eduardo Góes Neves, Museu de Arqueologia e Etnologia, Universidade de São Paulo, São Paulo, Brazil, edgneves@usp.br
Crystal N. McMichaels, Biological Sciences, Florida Institute of Technology, Melbourne, FL, cmcmicha@my.fit.edu
Christina Czarnecki, CSRC-University of New Hampshire, c.czarnecki@unh.edu
Stephen Hagen, Applied Geosolutions, LLC, steve.hagen@agsemail.com
Bobby H. Braswell, Applied Geosolutions, LLC, rob.braswell@gmail.com
Bruno Moraes, Museu de Arqueologia e Etnologia, Universidade de São Paulo, São Paulo, Brazil, bnomoraes@hotmail.com
Marco Raczka, Biological Sciences, Florida Institute of Technology, Melbourne, FL, mraczka2009@my.fit.edu

The pre-Columbian indigenous population estimates of the Amazon Basin lowlands are highly uncertain and the subject of considerable controversy. One of the archaeological sources used in reconstruction of Amazonian societies are Amazonian black earths (ABE) or in Portuguese, terra preta soils. The immense size of Amazonia, remoteness of many areas, forest vegetation, and lack of archaeological field surveys, make remote sensing beneficial to archaeological studies in this region. Remote sensing allows for comparison and analysis of vegetation across vast areas. Previous research has shown that hyperspectral image data can detect vegetation canopy chemistry differences, associated with soil nutrients and chemistry.

We conducted a preliminary analysis that indicates nine portions of the spectrum where the three ABE sites are completely separable from the three non-ABE sites. A discriminant function analysis using stepwise variable selection indicated that five bands were adequate in distinguishing between ABE and non-ABE sites. These five bands ranged between the 2000-2400 nm, indicating that canopy moisture is useful in remotely sensing terra preta. The wealth of site locations we are compiling from numerous sources provides a unique opportunity to develop algorithms for the classification of ABE and non-ABE sites. The large data volumes stemming from the 155 bands of hyperspectral data along with a great number of site sample points will produce a reliable relationship useful for statistical discrimination. The distribution and number of ABE sites provides information useful for both archaeological research and has consequences for the interpretation of Amazonian forest ecology. Knowledge of the disturbance history of the Amazonian forest provides a context and framework for the placement of all environmental research in the region.

Presentation Type:  Poster

Session:  Other   (Tue 11:30 AM)

Associated Project(s): 

  • Related Activity

Poster Location ID: 273

 


Close Window