Thread by @arjenvrielink

Dilemma: I can't stop myself from having mixes feelings about posts like these. On the one hand, great to advocate the power of #RemoteSensing and open datasets like @CopernicusEU. On the other hand, people expect this kind of quality for the South of Ghana as well. 1/

https://twitter.com/DescartesLabs/status/1133423492995661825

Most remotesensing marketing stuff is about:
* Moving gletsjers
* Wildfires
* Agriculture in the Noordoostpolder
* Crazy stuff in desert environments (no clouds)

So, great, but let's also show where the challenges are. And how we can help each other overcome those. 2/

Note to self: don't persist in being a bitter asshole, come up with some upsides and positive examples.

Self: well noted, working on it. /3

(Not really sure how these Twitter threads work ...)

Me, looking for positive examples but instead running into this thread from @vtcraghead proving my original point exactly. This is not what agricultural plots in developing countries look like.

https://twitter.com/vtcraghead/status/1133431924003880960?s=19

Hmm. I think this is called confirmation bias. Sorry. /4

Ok, some positive thoughts. What I think is going to be crucial for remote sensing based services to become widely adopted is strong support for lineage / provenance to provide transparancy, traceability and repeatability. Some really nice efforts are actively working on that /5

Projects working on transparancy, traceability and repeatability (standardisation, lineage):
* The stac project github.com/radiantearth/s…
* @open_EO openeo.org

and concepts like Analysis Ready Data, advocated by @opencholmes ao. /6

Efforts to reach smallholder farmers, in cooperation with big corp and including local context by collaborating with local scientists in Ghana medium.com/@wielaard/sate… (note: I work for @SatelligenceEO that published this work :-) ) /7