[Talk-GB] Tools to support solar panel mapping?

[SK53]

A few thoughts:

   1. The technologies used by OpenSolarMap by Christian Quest and others
   Etalab.could be applied to OS OpenLocal buildings with a suitable training
   set. The original French data used high quality imagery classified by cloud
   sourcing as to roof orientation (flat, E-W or N-S) and then applied this to
   the rest of France.
   2. The available data on installations is via the FIT scheme itself, but
   is only broken down by postcode district which is really too large for
   basic searching. A breakdown by postcode sector would be much more helpful
   & not likely to infringe any privacy aspects. I put a list of the top 25
   districts on the wiki.
   3. Additional information which could be used to identify candidates
   are: road orientation (E-W being best); housing age (available for most but
   not all MSOAs IIRC) with Victorian & pre WWI semis & detached houses being
   poor candidates, 30s & 70s council house terraces good ones; social housing
   (I have shape files for England based on NROSH data) as many HAs and
   at-length council housing arms have been very active in installing solar.
   4. I canvassed social housing experts on twitter for likely sites, again
   skimpily listed on the wiki.
   5. The new DG Vivid layers, at least near me, are much more recent and
   better for seeing rooftop solar installations.
   6. Scanning an area you know for rooftop solar installations is not too
   arduous, and could be done more systematically for smaller areas over the
   course of the quarter. I think Colm suggested mapping rooftop installations
   as nodes & I support this (at least in first instance). The huge benefit is
   that it often highlights other things which may be out-of-date or obviously
   in need of a survey, so it can fit well with 'local patch' mapping.
   7. I hope to soon blog about my analysis of Nottingham solar in terms of
   these external parameters (FIT installations, housing age, road
   orientation, social housing etc.).
   8. For ground solar, moisture index on Sentinel imagery can be useful to
   suggest candidates. We now also have access to a cloud free composite of
   2018 Sentinel in OSM editors.

Regards,

Jerry

On Thu, 23 May 2019 at 10:04, Jez Nicholson <jez.nicholson at gmail.com> wrote:

> Obviously we are talking about home/small-scale solar here. It could get
> quite involved, I'm sure that people are running whole businesses trying to
> analyse satellite imagery for this. Need to keep it simple and practical
> for this project, unless people have lots of time and energy to spare.
>
> An analysis (or link to an analysis) of the official stats could be
> useful. Exactly how did they make their estimates?
>
> Another idea: councils are making an effort to put panels on their
> properties. Could we FOI request them? or maybe someone has already done so.
>
>
>
> On Thu, May 23, 2019 at 8:59 AM Dan S <danstowell+osm at gmail.com> wrote:
>
>> Hi
>>
>> Related to the idea of solar panel mapping, I've had a request for
>> info about what sort of software tools might help support this work.
>> We might be using some of the familiar tools (e.g. streetcomplete,
>> openinframap, ... even tasking manager?).
>>
>> It'd be useful to have something like
>> completeness-by-postcode-district. Unlike Robert's postbox tools, we
>> don't have any official ID numbers for the items-to-map, we just have
>> some official stats (to be taken with a pinch of salt) about how many
>> are in each postcode district - but still, that could be a start.
>>
>> I'd also be interested in some tool that predicts where to look, which
>> might be based on analysing imagery, but perhaps more realistically
>> based on some mix of heuristics and official data.
>>
>> Any thoughts?
>>
>> Best
>> Dan
>>
>> _______________________________________________
>> Talk-GB mailing list
>> Talk-GB at openstreetmap.org
>> https://lists.openstreetmap.org/listinfo/talk-gb
>>
> _______________________________________________
> Talk-GB mailing list
> Talk-GB at openstreetmap.org
> https://lists.openstreetmap.org/listinfo/talk-gb
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/talk-gb/attachments/20190523/d69b1a88/attachment.html>