This week I attended an excellent talk given by Dr Jon Blower, CTO at IEA in Reading, entitled “Satellites, sensors and statistics – from environmental data to decisions” which was a meetup of the Society of Data Miners at the Royal Statistical Society in London. It is always good to attend events like this when the audience is predominately not experts in the field (Earth Observation) but are curious enough to make the effort to attend an evening meeting. To get a sense of the level of interest, 93 people attended/registered.
Some of the main advantages of Earth Observation Satellites (there are communications and positioning satellites as well) presented to the attendees included:
- Wide area capability – this could be used perhaps to complement UAV data
- Uniform coverage – the same sensor with the same configuration can cover many different places
- The increasing possibility of rapid response – targeting areas in remote locations. It’s possible to programme the satellite (though this is not a cheap option compared with already archived data)
- Continuity. This is a really exciting and incredibly important aspect of Earth Observation. Being able to have a long record of data and to be able to rely on data being acquired in the future is driving numerous business innovations. The Landsat series and the Copernicus mission put this as a very high priority. I am already looking forward to the forthcoming launch of Sentinel 2b due on 7th March on a Vega rocket
“It [Sentinel-2B] will join its twin, Sentinel-2A, which has been in orbit since June 2015. With each providing 290 km-wide coverage, the time it takes to image the globe will be cut in half to five days.”
One of the exciting things I heard as well was how much work is now being done with open source software. It reminded me of this tweet from 2015.
It's striking how much new earth imaging companies depend on open source & public data: Landsat, GDAL, landsat-util, etc. #SatSummit
— Robin Kraft (@robinkraft) November 9, 2015
It seems to me that for many years GIS/Geospatial people have been told/encouraged to move to Python – have a read of this this post, Yes, You Need to Code. There is a great piece of advice in that blog, written 2013:
“If you choose to get by with just using the GUI tools, you are doing yourself two disservices:
1.You are placing yourself at the mercy of others who can code to get around to building the customizations you need.
2. You are allowing your skills to erode by not using a significant amount of capability.”
Without any doubt in my mind today, Python is a key tool in your toolbox as an EO specialist (there are other tools, but…). It is incredibly exciting because organisations like the IEA are doing amazing things with free and open tools and open data, which was highlighted in the talk (some of the slides are here)