[HOT] Mapping task for seismic risk assessment

Joshua Campbell jcampbell.geo at gmail.com
Thu Oct 8 00:19:18 UTC 2015


John, no worries, we'll handle it. Appreciate your input.

But given that the buildings are not only shifted, but often incorrect, I
think it a better option to have them adjusted in real time (even with the
potential issues that raises). I've updated the task instructions to
indicate this.  I also don't expect there to be a massive number of
volunteers, so we'll keep an eye on the work as it progresses and determine
if we need to change the approach.

------------------
Joshua S Campbell
Geographer, PhD


On Wed, Oct 7, 2015 at 6:58 PM, john whelan <jwhelan0112 at gmail.com> wrote:

> So what we are saying then is the Bing imagery is fine but the many of the
> existing buildings do not align to the Bing imagery for one reason or
> another.  My concerns are new mappers will find this confusing and the
> effort in moving each building individually by individual mappers is to put
> it mildly a waste of resources.
>
> If someone who knows what they are doing pulls down the entire area into
> JOSM they could select each mappers buildings in turn and move them into
> place on the Bing imagery.  Yes I could do it but I have a batch of some
> twenty tiles to validate on a different project that I feel a greater
> commitment to.  Besides that one of my financial institutions made a
> mistake today which involved me going across town to sort it and and I'm
> still grumpy as you may have noticed.  Could someone else take the
> responsibility to line up the existing buildings please?
>
> Thanks
>
> Cheerio John
>
> On 7 October 2015 at 18:48, Blake Girardot <bgirardot at gmail.com> wrote:
>
>>
>> In looking at the osm gps traces and strava gps data, the current bing
>> imagery looks quite well aligned to me.
>>
>> cheers,
>> blake
>>
>> On 10/8/2015 12:20 AM, john whelan wrote:
>>
>>> I think the issue is how can we make this easy for new mappers.  You can
>>> select multiple buildings in JOSM by downloading the area then select
>>> buildings, then select an author within the selection then move all the
>>> buildings by that author at once.  It's not complicated just a pain to
>>> do if some of the buildings have been corrected already.
>>>
>>> Sometimes in OSM its less work to blast everything that is there
>>> checking the tags first to see its only buildings without additional
>>> tags then remap than it is to clean up each building.  However that
>>> approach is generally frowned on in OSM.
>>>
>>> May I suggest you freeze this project for the moment and get someone to
>>> nudge the existing buildings into place then open it up again.
>>>
>>> I'm not trying to pour cold water on the project but trying to head off
>>> a potential mess.
>>>
>>> Cheerio John
>>>
>>> On 7 October 2015 at 16:28, Joshua Campbell <jcampbell.geo at gmail.com
>>> <mailto:jcampbell.geo at gmail.com>> wrote:
>>>
>>>     John,
>>>
>>>     In reviewing the existing building data, you are correct it is
>>>     shifted. However, the data does not seem to be very accurate.
>>>
>>>     In this situation, I'm wondering it makes more sense to
>>>     double-check, correct, and update that existing data, rather than
>>>     try to bulk shift it (which I've never done). But I'm open to
>>>     suggestions on the best way forward.
>>>
>>>      >From the imagery perspective, the Bing Imagery analyzer indicates
>>>     the imagery is from May 2013. The DigitalGlobe archive shows a lot
>>>     of collection in that area in May 2013, but one scene in particular
>>>     (Catalog ID: 1030010022C8C400) is quite good. If this is the scene
>>>     in Bing, then I'm inclined to think the imagery is spatially correct
>>>     and serves as a good base to map from.
>>>
>>>     Regards,
>>>     Josh
>>>
>>>
>>>
>>>     On Tue, Oct 6, 2015 at 12:48 PM, john whelan <jwhelan0112 at gmail.com
>>>     <mailto:jwhelan0112 at gmail.com>> wrote:
>>>
>>>         Can some one double check the imagery, many buildings appear to
>>>         be already mapped but are displaced from the imagery.  Perhaps
>>>         some one could do a bulk move of the existing buildings?
>>>
>>>         http://tasks.hotosm.org/project/1235#task/187
>>>
>>>         Thanks John
>>>
>>>         On 6 October 2015 at 12:20, Joshua Campbell
>>>         <jcampbell.geo at gmail.com <mailto:jcampbell.geo at gmail.com>>
>>> wrote:
>>>
>>>             All,
>>>
>>>             Just wanted to announce the creation of a new mapping task
>>>             in the Kyrgyz Republic focused on seismic risk assessment
>>>             (http://tasks.hotosm.org/project/1235). The project is
>>>             funded by the Global Facility for Disaster Reduction and
>>>             Recovery, managed by World Bank and implemented by ARUP (and
>>>             partners) and includes seismic risk modeling to better
>>>             understand and manage earthquake risks.
>>>
>>>             The unique element to this task is that the resulting OSM
>>>             data will be directly integrated into a seismic risk model
>>>             which will be used by the Kyrgyz Republic Government for
>>>             decision making on risk reduction and preparedness.
>>>             Additionally the OSM building data will be assessed on the
>>>             ground and combined with a unique "streetview" type of
>>>             imagery data. We hope that some of you can help out with the
>>>             building mapping for this task, and help us test the
>>>             methodological advancement of using OSM data for risk
>>>             assessment.
>>>
>>>             Regards,
>>>             Josh
>>>
>>>             ------------------
>>>             Joshua S Campbell PhD
>>>             Senior Disaster Risk Management Specialist
>>>             Innovation Lab
>>>             World Bank GFDRR
>>>
>>>
>>>             _______________________________________________
>>>             HOT mailing list
>>>             HOT at openstreetmap.org <mailto:HOT at openstreetmap.org>
>>>             https://lists.openstreetmap.org/listinfo/hot
>>>
>>>
>>>
>>>
>>>
>>>
>>> _______________________________________________
>>> HOT mailing list
>>> HOT at openstreetmap.org
>>> https://lists.openstreetmap.org/listinfo/hot
>>>
>>>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/hot/attachments/20151007/1487b886/attachment.html>


More information about the HOT mailing list