[HOT] Including natural hazards in the HOSM data model?

Stéphane Henriod s at henriod.info
Wed Apr 4 21:03:11 BST 2012


Thank you all for your replies!

So let's try to keep the ball rolling, as I think we are now touching
something very interesting for HOT...

@Kate and Eric: "OpenStreetMap is usually used to map things that are
visible on the ground" and  "the 'average person' is unable to collect
relevant hazard information"

Let's leave the seismic hazard on the side for now (and other "exotic"
hazards, such as tsunami, flash flood or drought, as they have very
different onset mechanisms) and let's focus on earth mass movements
(landslides, rockfalls...) and floods.

>From my experience with communities in Central Asia, local people know very
well where the main hazard zones are located ("There's a minor landslide on
this foothill almost every year" or "My father told me that once there has
been a massive flood that destroyed 10 houses" or "This road is blocked by
avalanches every spring"...). The information is usually not very precise
(when? where exactly? how strong?...) but it often gives a pretty good idea
of the hazard situation, from which it is possible to derive an estimate of
he frequency and of the magnitude of a potential hazard.

So I would say that the "average person" is often able to delineate those
zones in his immediate living environment and that it is thus possible to
keep a bottom-up approach. Of course, it is usually necessary to engage
with local people in the field and it is rather hard to digitize anything
remotely. Moreover, such hazards often leave scars on the ground, which
make them "visible on the ground" (at least to some extent)

Of course, this is a history-based approach: we only find about hazards
that have occured during the last 1 or 2 generations. This makes the
approach imperfect, but should still capture most of the seasonal hazards.

@ Kate: "HOT's mission is to provide free geodata for use in response and
preparedness for disasters"

Then we probably agree that HOT should develop some standards for hazard
zones, as it has been done for post-crisis classes, right? The remaining
question is only whether those should be stored within OSM or not. Do we
all more or less agree on this statement?

Maybe the answer to this last question is given by Fran? "I'd be really
interested to see if Sahana would make a good place to
store this data". I have never worked with Sahana, so I would be happy if
someone with more experience could give his / her view? But if not, I'm
willing to investigate a little bit.

In this case, should we investigate the possibility to remove all of HOT
from OSM and to store the humanitarian data somewhere else? If I stick to
the definition "OpenStreetMap is usually used to map things that are
visible on the ground", which is obviously not the case for Search And
Rescue sectors, for example (at least not in a permanent way).

Once again, I'm really new to HOT and thus might not always be 100% to the
point, but I'm really excited about those discussions!

Should we start a wiki page to draft those ideas?

Stéphane
--
"Le mot progrès n'aura aucun sens tant qu'il y aura des enfants malheureux"
-- Albert Einstein

"A journey does not need reasons. Before long, it proves to be reason
enough in itself. One thinks that one is going to make a journey, yet soon
it is the journey that makes or unmakes you." -- Nicolas Bouvier

Photos de voyages, photos de montagne: http://www.henriod.info





On Wed, Apr 4, 2012 at 20:53, Kate Chapman <kate at maploser.com> wrote:

> I think what is important to think about are two points:
>
> 1. Which data is relevant for inclusion in OpenStreetMap
> 2. Which data is relevant for disaster preparedness and planning and
> should be open
>
> These two things aren't a one to one relation.  HOT's mission is to
> provide free geodata for use in response and preparedness for disasters.
>  Yes this usually translates to OpenStreetMap, but I think there are
> occasions where it would not.
>
> On Wed, Apr 4, 2012 at 11:43 AM, Eric Lovell <eric.j.lovell at gmail.com>wrote:
>
>> By assuming that the 'average person' is unable to collect relevant
>> hazard information, are we not claiming that only empirical top-down
>> information should be considered? Shouldn't place-based knowledge systems
>> be the most relevant systems for place-based hazards...that is "things that
>> are on the ground"? Isn't this counter intuitive to initiatives to
>> 'democratize' data?
>>
>> I have no input as to whether this type of information should be
>> incorporated into OSM. I think both Kate and Stéphane have valid points.
>> Just food for thought.
>>
>> Best,
>>
>> Eric
>>
>>
>> On Wed, Apr 4, 2012 at 11:18 AM, Kate Chapman <kate at maploser.com> wrote:
>>
>>> Hi Stéphane,
>>>
>>> The issue I see is more that it will depend on the type of hazard data.
>>>  OpenStreetMap is usually used to map things that are visible on the
>>> ground.  I see it being difficult for some types of hazards to be collected
>>> by the average person (not to say for some types it isn't easier).  I'm not
>>> saying it is less relevant to humanitarian actors, I'm saying that it
>>> perhaps is better in another store and to be compared with OpenStreetMap
>>> data.
>>>
>>> People combine data with OSM data all the time.  Everything isn't either
>>> an in or out proposition.  I think it is important to have relevant and
>>> updatable information in OSM and be able to utilize it with other
>>> information that might be less practical for us to collect.
>>>
>>> -Kate
>>>
>>>
>>> On Wed, Apr 4, 2012 at 10:35 AM, Stéphane Henriod <s at henriod.info>wrote:
>>>
>>>> Hi Kate,
>>>>
>>>> I am browsing through inasafe website and it seems indeed pretty
>>>> interesting and relevant for the issues I raised in my first email.
>>>>
>>>> However, I believe that it is complementary rather than conflicting:
>>>> what I would aim at is the possibility to collect hazard data in an
>>>> OSM-compliant format that could, in turn, be used for risk analysis, for
>>>> example with InaSAFE.
>>>>
>>>> I see your point that hazard data (flood-prone areas, recurrent
>>>> landslides, seismic zones...) might need to be located out of the main OSM
>>>> DB, but I would be interested to read whether this opinion is mostly shared
>>>> by other HOT members? What I mean is: is the hazard data really less
>>>> relevant to the humanitarian and to the overall communities than the
>>>> blocked roads, the IDPs locations or the Search and Rescue zones? IMHO,
>>>> they would fall in the same category, so either everything in the main OSM
>>>> DB, or everything out of it. Does that sound silly?
>>>>
>>>> Thanks a lot for replying to me, and looking forward to engage the
>>>> discussion with any other member that would feel interested!
>>>>
>>>>
>>>> Stéphane
>>>> --
>>>> "Le mot progrès n'aura aucun sens tant qu'il y aura des enfants
>>>> malheureux" -- Albert Einstein
>>>>
>>>> "A journey does not need reasons. Before long, it proves to be reason
>>>> enough in itself. One thinks that one is going to make a journey, yet soon
>>>> it is the journey that makes or unmakes you." -- Nicolas Bouvier
>>>>
>>>> Photos de voyages, photos de montagne: http://www.henriod.info
>>>>
>>>> Skype: [image: Skype name: marmotte_la_gueuse]
>>>> Tajik mobile phone: +992 934 62 46 62
>>>>
>>>>
>>>>
>>>> On Wed, Apr 4, 2012 at 17:06, Kate Chapman <kate at maploser.com> wrote:
>>>>
>>>>> Hello Stéphane,
>>>>>
>>>>> I don't think it would really make sense to add many of those types of
>>>>> hazards to OpenStreetMap.  We've collected information such as areas
>>>>> prone to flooding before but the actual hazard models should probably
>>>>> stay separate.  They can be combined with OpenStreetMap information
>>>>> however to create impact models.
>>>>>
>>>>> This is something currently being done by AIFDR and GFDRR with
>>>>> InaSAFE: https://github.com/AIFDR/inasafe
>>>>>
>>>>> Best,
>>>>>
>>>>> -Kate
>>>>>
>>>>> On Wed, Apr 4, 2012 at 7:14 AM, Stéphane Henriod <s at henriod.info>
>>>>> wrote:
>>>>> > Hello,
>>>>> >
>>>>> > I am relatively new to OSM and completely new to Humanitarian OSM but
>>>>> > extremely enthusiastic about both projects and would like to
>>>>> contribute in a
>>>>> > way that I will describe below, and for which I would appreciate your
>>>>> > feedback and ideas.
>>>>> >
>>>>> > Most humanitarian actors agree that the best response to an
>>>>> emergency is
>>>>> > preparedness, not only in terms of behaviors but also data. What
>>>>> strikes me
>>>>> > when I look at HOSM data model is that I don't see anything for
>>>>> natural
>>>>> > hazard delineation, for example. When responding to a crisis,
>>>>> humanitarian
>>>>> > actors might want to know in advance where to expect landslides,
>>>>> avalanches
>>>>> > and floods areas (for their own safety but also to "predict" where
>>>>> roads
>>>>> > might be blocked or where facilities might have been affected).
>>>>> Seismic
>>>>> > zonation is something that can be directly included in OSM; oldish
>>>>> data is
>>>>> > freely available from the GSHAP project, while the GlobalEarthquake
>>>>> Model
>>>>> > will provide with a more modern version in a few years.
>>>>> >
>>>>> > My question to the active and experienced members of the community
>>>>> is thus
>>>>> > the following: would you see any relevance to develop classes for
>>>>> natural
>>>>> > (and man-made?) hazards to be included in HOSM framework?
>>>>> >
>>>>> > As we can not expect the communities to have the technical knowledge
>>>>> of
>>>>> > hazard scientists, the ontology would have to remain quite simple,
>>>>> which is
>>>>> > probably ok for humanitarian actors (who cares whether it's a
>>>>> landslide, a
>>>>> > rockfall or a debris flow? a general category "earth mass movement"
>>>>> might be
>>>>> > sufficient for our purpose). Also, a distinction should be made
>>>>> between
>>>>> > "usual" events (those that occur relatively frequently but that
>>>>> might or
>>>>> > might not be currently triggered) and actual events (that have been
>>>>> actually
>>>>> > triggered).
>>>>> >
>>>>> > Many challenges ahead but before diving into it, I would like to
>>>>> check with
>>>>> > you if this idea has already been discussed and if it makes sense to
>>>>> pursue
>>>>> > it.
>>>>> >
>>>>> > Thanks a lot in advance and look forward to reading your comments,
>>>>> >
>>>>> > Stéphane
>>>>> >
>>>>> >
>>>>> > --
>>>>> > "Le mot progrès n'aura aucun sens tant qu'il y aura des enfants
>>>>> malheureux"
>>>>> > -- Albert Einstein
>>>>> >
>>>>> > "A journey does not need reasons. Before long, it proves to be
>>>>> reason enough
>>>>> > in itself. One thinks that one is going to make a journey, yet soon
>>>>> it is
>>>>> > the journey that makes or unmakes you." -- Nicolas Bouvier
>>>>> >
>>>>> > Photos de voyages, photos de montagne: http://www.henriod.info
>>>>> >
>>>>> >
>>>>> > _______________________________________________
>>>>> > HOT mailing list
>>>>> > HOT at openstreetmap.org
>>>>> > http://lists.openstreetmap.org/listinfo/hot
>>>>> >
>>>>>
>>>>
>>>>
>>>> _______________________________________________
>>>> HOT mailing list
>>>> HOT at openstreetmap.org
>>>> http://lists.openstreetmap.org/listinfo/hot
>>>>
>>>>
>>>
>>> _______________________________________________
>>> HOT mailing list
>>> HOT at openstreetmap.org
>>> http://lists.openstreetmap.org/listinfo/hot
>>>
>>>
>>
>>
>>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/hot/attachments/20120404/36fefe9c/attachment-0001.html>


More information about the HOT mailing list