Below I commented on a 1998 USGS report on the availability of oil in the ANWR "1002 Area". This report superceded a previous USGS report released in 1987. Both of these were based on the same seismic survey data, taken over the winters of '83-84 and '84-85 (the 1998 report also included data from nearby wells). These were 2-D seismic surveys; since performing such a survey in a protected region requires an act of Congress (for reasons illustrated below), this is the ONLY such data ever generated from the region, and that is likely to remain the case.

A seismic survey works as follows: the survey vehicle is a 30-ton behemoth with a big metal plate attached underneath. At regular intervals it stops, sets the plate down, rests its weight on it, and sends out vibrations. These are picked up by seismic sensing equipment nearby; thus a seismic map can be generated.

2D seismic surveys and 3D seismic surveys both operate on similar data sets; it's the post-processing of the data that represent the technological leap in the latter. The only difference in a 3D survey is that the area needs to be much more finely sampled, with a grid density an order of magnitude higher than 2D surveys.

The geology of ANWR is such that the oil is distributed in many hundreds of relatively small pockets (in contrast to the nearby Prudhoe Bay, where most of the oil lies in a single giant field). Thus, to prevent the expense of drilling many, many dry wells, a comprehensive and accurate survey is in order, i.e. 3D or possibly 4D (time-lapsed) seismic surveys.

Here is a US Fish and Wildlife Survey image showing the relative density of a 3D survey to the 2D survey performed. And below is a dyad of images showing the effects of some of the survey vehicles from the 1984-5 survey. Scarring like that depicted doesn't happen everywhere, but it's safe to say that the much denser survey (especially if done repeatedly) would have a strong impact on the surprisingly delicate arctic tundra.

So, irrespective of the damage done by actual drilling, gravel roads, pipelines, etc., the impact done simply by exploration would be considerable.

The FWS page on the subject of ANWR, by the way, is very good and worth reading, here.