Last April I used Landsat 7 images to document the impact of the January 2013 in the Ica-Nazca sector of the Peruvian coastal desert. Despite claims by the rally organisers that the route was jointly devised in cooperation with the Peruvian Ministry of Culture, sequential satellite images clearly showed that the rally had crossed a large geoglyph. Worries about archaeological heritage in the region rose again when a Dakar Challenge rally was announced for early October 2013. Again, this rally was to take place in the desert between the city of Ica and the Pacific Ocean. Again, not much that I could do about it but wait until it was over, hope for the best and then analyse sequential satellite images to see where changes had taken place – and if they had again affected archaeological sites.
This time I used images acquired by the new Landsat 8 satellite. Spatial resolution of the panchromatic band is the same as for Landsat 7 (15 m), but it’s 16 bit data now, so image quality has been improved. More importantly, while Landsat 7 images have since 2003 been affected by black stripes due to the failure of the scan line corrector, Landsat 8 now provides full coverage. That issue had not been a problem in the case of the geoglyphs near Santa Cruz which I looked at earlier because they are close to the centre of the scene (and the striping increases from scene centre towards the margins).
The image processing was similar to what I did last time: get images before and after the dates of the rally, normalise one image to the other to get rid of brightness changes due to changing illumination, subtract the earlier image from the newer one and then find areas where changes had occured. The first result of this is a map of tracks left by offroad vehicles between 13 September 2013 (before the rally) and 16 November 2013 (after the rally). Also shown on this map are the locations of known geoglyphs as listed by Asociación Maria Reiche. At the first sight, most offroad traffic has occured at a save distance from the geoglyphs. But at some sites, I had to look closer.
I therefore loaded the mapped tracks into Google Earth to see whether they cross geoglyphs. And unfortunately, yes, they do. In one case, a large geoglyph was crossed by the rally. Here, the impact of the rally on the geoglyph is very clear as shown by the changes in the Landsat images.
In a second case, assessing the actual impact of the Dakar rally on geoglyphs is more difficult. Here, fresh offroad vehicle tracks are only visible in some areas. However, extrapolating the tracks into terrain where the vehicles did not churn up as much dust (and where the tracks are therefore not recognisable in the relatively coarse resolution Landsat images) shows that damage to other geoglyphs is very likely.
Yet again, a Dakar rally has caused irreversible damage to known archaeological sites. And these are only the large, conspicuous site visible in satellite images. Analysing Landsat images to document damages to archaeological heritage requires that we know where the sites are. This is relatively easy for sites like large geoglyphs, but finding smaller sites requires time-intensive field work. From the limited time that I have spent walking in this desert I know that the number of archaeological sites – shell middens, windbreaks, ancient paths etc. – is surprisingly high. Treating this desert as if it consisted only of sand and rocks by thoughtlessly using it as a playground for offroad vehiles is irresponsible.
The January 2014 Dakar rally did not take place in Peru. But they are planning to come back.