It would be really great to be able to deploy your own sensor network instead of pleading with busy and underfunded governmental agencies to help you out.
I'm very much interested in monitoring air pollutants. I live in a place in Germany where BASF built "the worlds largest chemical plant" right next to the centers of two cities with about half a million residents combined [2]. And of course, you can smell that they pollute the air with something. Yet there is no information whatsoever on which chemicals they dump into the air and what the concentration is. The usual air quality measurements cover O3, SO2, NO2 and PM10 [3]. It's probably easy to monitor those yourself, just buy a sensor. But how do you monitor something that you can smell, you don't know what it is? Mass spectrometer?
[0] http://www.youtube.com/watch?v=UyGjdloVPKQ
[1] http://blog.safecast.org/
[2] http://www.openstreetmap.org/?lat=49.4947&lon=8.4424&zoom=14&layers=MThe aggregate data could be really interesting.
[1] http://www.kickstarter.com/projects/robocat/thermodo-the-tin...
My startup builds big sensor networks - but the sensor side of things is where the real innovation is. Microfluidics, 'labs on a chip' in CMOS, infrared spectroscopy, etc...
Sadly it's worlds different than using thermocouples or reading i2c type buses.
Unfortunately much of these advances cited above are academic efforts. How this translates to a product usable 'in the field' is perhaps more clear in the medical device world than in environmental monitoring.
I'd like to believe one could 'crowd-source' pollution monitoring. Open source sensors, drive costs down, collect data, throw it atop an OpenStreetMap layer and somehow the populus as a whole could make better informed decisions in resource consumption/politics/whatever.
Yet these types of studies (presently) are best carried out by governments and academic researchers for a number of reasons (land rights, use of 'mixed use' satellites, calibration, etc). Take water pollution for instance. The USGS builds 'The National Map' [1] which has a data layer for water runoff [2] (rivers,streams,etc), pollution sources [3], and whatever data sources are hidden in EPA, private mining/gas industires, Army Corp of Engineers, etc databases. The GIS work to make sense of even existing data is daunting. Add to it the polical winds of Fiscal Sequestration and 'less-regulation' and things look grim.
That said, microfluidics and biomarkers are moving more ino the Hacker realm and out of the cleanroom. And ultimately it's how well the Hacker community interacts with academia/goverment...