Re: [HOT] landslides and imagery

[john o'l]

Well I think I'm rapidly approaching the end of the window I had to work
with this and figure out how to add it to osm, if anyone else wants to try,
go for it.  In the meantime, I just figured out how to turn it into a kml,
it contains sites in addition to those mapped by British Geological Survey,
Durham University, ICIMOD, NASA-JPL, Univ. of Arizona -- though one
coincides: theirs: 20150508_hazard_db:FID 120 this eq_kml_shp:id -80 and
eq_kml_shp:id -81. Their db is available at:
https://data.hdx.rwlabs.org/dataset/lands

This kmz here (rougher than I'd like but it will have to do... )  All the
best, John


On Mon, May 11, 2015 at 8:59 AM, john o'l  wrote:

> Also linking a website with links to a report on the Oso landslide as it
> is somewhat familiar to me
> http://www.geerassociation.org/GEER_Post%20EQ%20Reports/Oso_WA_2014/index.html
> For our purposes, some of the key points made are:  the initial phase of
> the slide (most destructive/fastest and most distant runout) -- was an old
> slide that had remobilized. 200 vertical meters of material which covered a
> horizontal distance of 1000 meters.  So existing landslides can remain very
> dangerous for a considerable period of time.  The 1 km runout, (or
> "deposit" keeping with the tagging scheme I proposed  "area = deposit") was
> not extraordinary based on volume of material, however it was extraordinary
> relative to most peoples' perceptions -- I doubt anyone seriously
> considered that a hillside that low and distant could be an active risk to
> an area as far away as the main highway corridor.
>
> After mapping some more landslides to my local machine but yet to upload
> any to OSM, I'd like to update my proposed OSM tagging scheme:
>
> hazard_type = landslide
> hazard_prone = yes
> area = scarp OR deposit
> damage:event = nepal_earthquake_2015 OR pre_nepal_earthquake_2015
> barrier = scarp OR deposit
> source = DigitalGlobe, 2015-05 OR other as appropriate
> landuse = brownfield
>
> Cheers,
>
> John
>
> On Mon, May 11, 2015 at 8:00 AM, john o'l  wrote:
>
>> Thanks Prabhas,
>>
>> Very interesting! Yesterday I was directed to the Earthquakes Without
>> Frontiers blog http://ewf.nerc.ac.uk/blog/ and a map linked from their
>> May 8 post
>> http://ewf.nerc.ac.uk/wp-content/uploads/2015/05/Landslide_Update_2_08052015_SMALL.jpg,
>> apparently higher resolution is also available..
>>
>> Cheers,
>> John
>>
>> On Mon, May 11, 2015 at 7:10 AM, Prabhas Pokharel <
>> prabhas.pokha...@gmail.com> wrote:
>>
>>> Another email to add to the list for those interested in doing landslide
>>> mapping:
>>> We at KLL were forwarded this landslide risk assessment layer:
>>> https://www.google.com/maps/d/u/0/viewer?mid=z6HUO2aILzmQ.kGtOdlu45GXY&usp=sharing
>>> which comes from here:
>>> https://sites.google.com/a/umich.edu/nepalearthquake/landslide-maps
>>>
>>> It may help those of us interested in finding lanslides have some areas
>>> of high risk where they could start looking.
>>>
>>> cheers,
>>> Prabhas
>>>
>>> On Sun, May 10, 2015 at 11:50 AM, Springfield Harrison <
>>> stellar...@gmail.com> wrote:
>>>
 Hello John,

 With reference to your moving boulder, just wondering if that could be
 in fact moving, i.e., not an image based coordinate shift as such.  I'm
 just thinking that with aftershocks and general instability, many of these
 new features are still sorting themselves out and traveling downhill.

 Can DG or Bing make stereo pairs available?  Likely a long shot, but
 thought I would ask.

 Cheers . . . . . . . . Spring Harrison


 At 09-05-2015 17:49 Saturday, john o'l wrote:

> I've been focusing on landslides and have located several score that
> appear recent. Of these, a few are pre-quake and appear relatively 
> stable,
> some are pre-quake but appear reactivated and many appear to have been
> associated with the quake and/or aftershocks. I've mapped several dozen of
> these so far. In my next email, I'll cover why you won't find them in
> osm... yet. For this one, I'd like to stick to post quake imagery and some
> of its quirks.Â
>
> There is an inhabited hillside that had numerous landslides, some
> predate the quake, but most are presumably related. So far I've mapped
> about half of them, those that are largest or appear to threaten buildings
> and pathways. There is Digital Globe imagery available from May 3 and May
> 8. It looks like QGIS easily operates with more than one coordinate system
> at a time. The center of a large boulder in the May 3 imagery (Longitude,
> Latitude; WGS84 EPSG:3857 x,y) is at
> 85.85659,27.83609;9557511.789,3228324.329, in the May 8 imagery it is at
> 85.85669,27.83656;9557522.728,3228382.865. Mind you, this is not a
> complaint, rather it is a concrete example of the variability with this
> recent imagery. Â
>
> A more extreme example is a slide that appea

Re: [HOT] landslides and imagery

[john o'l]

Also linking a website with links to a report on the Oso landslide as it is
somewhat familiar to me
http://www.geerassociation.org/GEER_Post%20EQ%20Reports/Oso_WA_2014/index.html
For our purposes, some of the key points made are:  the initial phase of
the slide (most destructive/fastest and most distant runout) -- was an old
slide that had remobilized. 200 vertical meters of material which covered a
horizontal distance of 1000 meters.  So existing landslides can remain very
dangerous for a considerable period of time.  The 1 km runout, (or
"deposit" keeping with the tagging scheme I proposed  "area = deposit") was
not extraordinary based on volume of material, however it was extraordinary
relative to most peoples' perceptions -- I doubt anyone seriously
considered that a hillside that low and distant could be an active risk to
an area as far away as the main highway corridor.

After mapping some more landslides to my local machine but yet to upload
any to OSM, I'd like to update my proposed OSM tagging scheme:

hazard_type = landslide
hazard_prone = yes
area = scarp OR deposit
damage:event = nepal_earthquake_2015 OR pre_nepal_earthquake_2015
barrier = scarp OR deposit
source = DigitalGlobe, 2015-05 OR other as appropriate
landuse = brownfield

Cheers,

John

On Mon, May 11, 2015 at 8:00 AM, john o'l  wrote:

> Thanks Prabhas,
>
> Very interesting! Yesterday I was directed to the Earthquakes Without
> Frontiers blog http://ewf.nerc.ac.uk/blog/ and a map linked from their
> May 8 post
> http://ewf.nerc.ac.uk/wp-content/uploads/2015/05/Landslide_Update_2_08052015_SMALL.jpg,
> apparently higher resolution is also available..
>
> Cheers,
> John
>
> On Mon, May 11, 2015 at 7:10 AM, Prabhas Pokharel <
> prabhas.pokha...@gmail.com> wrote:
>
>> Another email to add to the list for those interested in doing landslide
>> mapping:
>> We at KLL were forwarded this landslide risk assessment layer:
>> https://www.google.com/maps/d/u/0/viewer?mid=z6HUO2aILzmQ.kGtOdlu45GXY&usp=sharing
>> which comes from here:
>> https://sites.google.com/a/umich.edu/nepalearthquake/landslide-maps
>>
>> It may help those of us interested in finding lanslides have some areas
>> of high risk where they could start looking.
>>
>> cheers,
>> Prabhas
>>
>> On Sun, May 10, 2015 at 11:50 AM, Springfield Harrison <
>> stellar...@gmail.com> wrote:
>>
>>> Hello John,
>>>
>>> With reference to your moving boulder, just wondering if that could be
>>> in fact moving, i.e., not an image based coordinate shift as such.  I'm
>>> just thinking that with aftershocks and general instability, many of these
>>> new features are still sorting themselves out and traveling downhill.
>>>
>>> Can DG or Bing make stereo pairs available?  Likely a long shot, but
>>> thought I would ask.
>>>
>>> Cheers . . . . . . . . Spring Harrison
>>>
>>>
>>> At 09-05-2015 17:49 Saturday, john o'l wrote:
>>>
 I've been focusing on landslides and have located several score that
 appear recent. Of these, a few are pre-quake and appear relatively stable,
 some are pre-quake but appear reactivated and many appear to have been
 associated with the quake and/or aftershocks. I've mapped several dozen of
 these so far. In my next email, I'll cover why you won't find them in
 osm... yet. For this one, I'd like to stick to post quake imagery and some
 of its quirks.Â

 There is an inhabited hillside that had numerous landslides, some
 predate the quake, but most are presumably related. So far I've mapped
 about half of them, those that are largest or appear to threaten buildings
 and pathways. There is Digital Globe imagery available from May 3 and May
 8. It looks like QGIS easily operates with more than one coordinate system
 at a time. The center of a large boulder in the May 3 imagery (Longitude,
 Latitude; WGS84 EPSG:3857 x,y) is at
 85.85659,27.83609;9557511.789,3228324.329, in the May 8 imagery it is at
 85.85669,27.83656;9557522.728,3228382.865. Mind you, this is not a
 complaint, rather it is a concrete example of the variability with this
 recent imagery. Â

 A more extreme example is a slide that appeared to be partially
 blocking a stream in the May 3 imagery 85.90258,27.87818;9562631.312,
 3233623.303; -- it was completely obscured by a hillside in the May 8
 imagery (probably taken from a more northerly or northwesterly vantage
 point.)Â

 Downslope (westward) from a likely reactivated slide located at
 85.81987,27.90810;9553423.739,3237391.771 Â is a remote area that
 appears very hard hit. The May 8th imagery is mostly clouds, but the May
 3rd imagery shows a blue rooftop at
 85.80644,27.90818;9551929.301,3237402.414, it looks like there are several
 large boulders in the immediate area and there is not much left to tell
 there were more than 20 buildings nearby. While the boulders may have
 contributed, at the moment I think

Re: [HOT] landslides and imagery

[john o'l]

Thanks Prabhas,

Very interesting! Yesterday I was directed to the Earthquakes Without
Frontiers blog http://ewf.nerc.ac.uk/blog/ and a map linked from their May
8 post
http://ewf.nerc.ac.uk/wp-content/uploads/2015/05/Landslide_Update_2_08052015_SMALL.jpg,
apparently higher resolution is also available..

Cheers,
John

On Mon, May 11, 2015 at 7:10 AM, Prabhas Pokharel <
prabhas.pokha...@gmail.com> wrote:

> Another email to add to the list for those interested in doing landslide
> mapping:
> We at KLL were forwarded this landslide risk assessment layer:
> https://www.google.com/maps/d/u/0/viewer?mid=z6HUO2aILzmQ.kGtOdlu45GXY&usp=sharing
> which comes from here:
> https://sites.google.com/a/umich.edu/nepalearthquake/landslide-maps
>
> It may help those of us interested in finding lanslides have some areas of
> high risk where they could start looking.
>
> cheers,
> Prabhas
>
> On Sun, May 10, 2015 at 11:50 AM, Springfield Harrison <
> stellar...@gmail.com> wrote:
>
>> Hello John,
>>
>> With reference to your moving boulder, just wondering if that could be in
>> fact moving, i.e., not an image based coordinate shift as such.  I'm just
>> thinking that with aftershocks and general instability, many of these new
>> features are still sorting themselves out and traveling downhill.
>>
>> Can DG or Bing make stereo pairs available?  Likely a long shot, but
>> thought I would ask.
>>
>> Cheers . . . . . . . . Spring Harrison
>>
>>
>> At 09-05-2015 17:49 Saturday, john o'l wrote:
>>
>>> I've been focusing on landslides and have located several score that
>>> appear recent. Of these, a few are pre-quake and appear relatively stable,
>>> some are pre-quake but appear reactivated and many appear to have been
>>> associated with the quake and/or aftershocks. I've mapped several dozen of
>>> these so far. In my next email, I'll cover why you won't find them in
>>> osm... yet. For this one, I'd like to stick to post quake imagery and some
>>> of its quirks.Â
>>>
>>> There is an inhabited hillside that had numerous landslides, some
>>> predate the quake, but most are presumably related. So far I've mapped
>>> about half of them, those that are largest or appear to threaten buildings
>>> and pathways. There is Digital Globe imagery available from May 3 and May
>>> 8. It looks like QGIS easily operates with more than one coordinate system
>>> at a time. The center of a large boulder in the May 3 imagery (Longitude,
>>> Latitude; WGS84 EPSG:3857 x,y) is at
>>> 85.85659,27.83609;9557511.789,3228324.329, in the May 8 imagery it is at
>>> 85.85669,27.83656;9557522.728,3228382.865. Mind you, this is not a
>>> complaint, rather it is a concrete example of the variability with this
>>> recent imagery. Â
>>>
>>> A more extreme example is a slide that appeared to be partially blocking
>>> a stream in the May 3 imagery 85.90258,27.87818;9562631.312,3233623.303;
>>> -- it was completely obscured by a hillside in the May 8 imagery (probably
>>> taken from a more northerly or northwesterly vantage point.)Â
>>>
>>> Downslope (westward) from a likely reactivated slide located at
>>> 85.81987,27.90810;9553423.739,3237391.771 Â is a remote area that
>>> appears very hard hit. The May 8th imagery is mostly clouds, but the May
>>> 3rd imagery shows a blue rooftop at
>>> 85.80644,27.90818;9551929.301,3237402.414, it looks like there are several
>>> large boulders in the immediate area and there is not much left to tell
>>> there were more than 20 buildings nearby. While the boulders may have
>>> contributed, at the moment I think it is probable that the shaking itself
>>> was mostly responsible for the extreme level of destruction.Â
>>>
>>> One advantage of different acquisition angles is that some features may
>>> be discernible on slopes that don't ordinarily show up very well.Â
>>> Â
>>> Question to the HOT folks -- is there a way to specify the date of DG
>>> imagery we access through the proxy server?, Some of the May 8 imagery is
>>> starting to come up over the May 3 imagery without me telling it to. Â
>>>
>>> Best regards,
>>>
>>> John
>>> ___ HOT mailing list
>>> HOT@openstreetmap.org https://lists.openstreetmap.org/listinfo/hot
>>>
>>
>>
>> ___
>> HOT mailing list
>> HOT@openstreetmap.org
>> https://lists.openstreetmap.org/listinfo/hot
>>
>
>
>
> --
> Prabhas Pokharel
> http://prabhasp.com
> twitter/skype/facebook/whatever: prabhasp
>
___
HOT mailing list
HOT@openstreetmap.org
https://lists.openstreetmap.org/listinfo/hot

Re: [HOT] landslides and imagery

[Prabhas Pokharel]

Another email to add to the list for those interested in doing landslide
mapping:
We at KLL were forwarded this landslide risk assessment layer:
https://www.google.com/maps/d/u/0/viewer?mid=z6HUO2aILzmQ.kGtOdlu45GXY&usp=sharing
which comes from here:
https://sites.google.com/a/umich.edu/nepalearthquake/landslide-maps

It may help those of us interested in finding lanslides have some areas of
high risk where they could start looking.

cheers,
Prabhas

On Sun, May 10, 2015 at 11:50 AM, Springfield Harrison  wrote:

> Hello John,
>
> With reference to your moving boulder, just wondering if that could be in
> fact moving, i.e., not an image based coordinate shift as such.  I'm just
> thinking that with aftershocks and general instability, many of these new
> features are still sorting themselves out and traveling downhill.
>
> Can DG or Bing make stereo pairs available?  Likely a long shot, but
> thought I would ask.
>
> Cheers . . . . . . . . Spring Harrison
>
>
> At 09-05-2015 17:49 Saturday, john o'l wrote:
>
>> I've been focusing on landslides and have located several score that
>> appear recent. Of these, a few are pre-quake and appear relatively stable,
>> some are pre-quake but appear reactivated and many appear to have been
>> associated with the quake and/or aftershocks. I've mapped several dozen of
>> these so far. In my next email, I'll cover why you won't find them in
>> osm... yet. For this one, I'd like to stick to post quake imagery and some
>> of its quirks.Â
>>
>> There is an inhabited hillside that had numerous landslides, some predate
>> the quake, but most are presumably related. So far I've mapped about half
>> of them, those that are largest or appear to threaten buildings and
>> pathways. There is Digital Globe imagery available from May 3 and May 8. It
>> looks like QGIS easily operates with more than one coordinate system at a
>> time. The center of a large boulder in the May 3 imagery (Longitude,
>> Latitude; WGS84 EPSG:3857 x,y) is at
>> 85.85659,27.83609;9557511.789,3228324.329, in the May 8 imagery it is at
>> 85.85669,27.83656;9557522.728,3228382.865. Mind you, this is not a
>> complaint, rather it is a concrete example of the variability with this
>> recent imagery. Â
>>
>> A more extreme example is a slide that appeared to be partially blocking
>> a stream in the May 3 imagery 85.90258,27.87818;9562631.312,3233623.303;
>> -- it was completely obscured by a hillside in the May 8 imagery (probably
>> taken from a more northerly or northwesterly vantage point.)Â
>>
>> Downslope (westward) from a likely reactivated slide located at
>> 85.81987,27.90810;9553423.739,3237391.771 Â is a remote area that
>> appears very hard hit. The May 8th imagery is mostly clouds, but the May
>> 3rd imagery shows a blue rooftop at
>> 85.80644,27.90818;9551929.301,3237402.414, it looks like there are several
>> large boulders in the immediate area and there is not much left to tell
>> there were more than 20 buildings nearby. While the boulders may have
>> contributed, at the moment I think it is probable that the shaking itself
>> was mostly responsible for the extreme level of destruction.Â
>>
>> One advantage of different acquisition angles is that some features may
>> be discernible on slopes that don't ordinarily show up very well.Â
>> Â
>> Question to the HOT folks -- is there a way to specify the date of DG
>> imagery we access through the proxy server?, Some of the May 8 imagery is
>> starting to come up over the May 3 imagery without me telling it to. Â
>>
>> Best regards,
>>
>> John
>> ___ HOT mailing list
>> HOT@openstreetmap.org https://lists.openstreetmap.org/listinfo/hot
>>
>
>
> ___
> HOT mailing list
> HOT@openstreetmap.org
> https://lists.openstreetmap.org/listinfo/hot
>



-- 
Prabhas Pokharel
http://prabhasp.com
twitter/skype/facebook/whatever: prabhasp
___
HOT mailing list
HOT@openstreetmap.org
https://lists.openstreetmap.org/listinfo/hot

Re: [HOT] landslides and imagery

[Springfield Harrison]

Hello John,

With reference to your moving boulder, just 
wondering if that could be in fact moving, i.e., 
not an image based coordinate shift as such.  I'm 
just thinking that with aftershocks and general 
instability, many of these new features are still 
sorting themselves out and traveling downhill.


Can DG or Bing make stereo pairs 
available?  Likely a long shot, but thought I would ask.


Cheers . . . . . . . . Spring Harrison


At 09-05-2015 17:49 Saturday, john o'l wrote:
I've been focusing on landslides and have 
located several score that appear recent. Of 
these, a few are pre-quake and appear relatively 
stable, some are pre-quake but appear 
reactivated and many appear to have been 
associated with the quake and/or aftershocks. 
I've mapped several dozen of these so far. In my 
next email, I'll cover why you won't find them 
in osm... yet. For this one, I'd like to stick 
to post quake imagery and some of its quirks.Â


There is an inhabited hillside that had numerous 
landslides, some predate the quake, but most are 
presumably related. So far I've mapped about 
half of them, those that are largest or appear 
to threaten buildings and pathways. There is 
Digital Globe imagery available from May 3 and 
May 8. It looks like QGIS easily operates with 
more than one coordinate system at a time. The 
center of a large boulder in the May 3 imagery 
(Longitude, Latitude; WGS84 EPSG:3857 x,y) is at 
85.85659,27.83609;9557511.789,3228324.329, in 
the May 8 imagery it is at 
85.85669,27.83656;9557522.728,3228382.865. Mind 
you, this is not a complaint, rather it is a 
concrete example of the variability with this recent imagery. Â


A more extreme example is a slide that appeared 
to be partially blocking a stream in the May 3 
imagery 
85.90258,27.87818;9562631.312,3233623.303; -- it 
was completely obscured by a hillside in the May 
8 imagery (probably taken from a more northerly 
or northwesterly vantage point.)Â


Downslope (westward) from a likely reactivated 
slide located at 
85.81987,27.90810;9553423.739,3237391.771 Â is a 
remote area that appears very hard hit. The May 
8th imagery is mostly clouds, but the May 3rd 
imagery shows a blue rooftop at 
85.80644,27.90818;9551929.301,3237402.414, it 
looks like there are several large boulders in 
the immediate area and there is not much left to 
tell there were more than 20 buildings nearby. 
While the boulders may have contributed, at the 
moment I think it is probable that the shaking 
itself was mostly responsible for the extreme level of destruction.Â


One advantage of different acquisition angles is 
that some features may be discernible on slopes 
that don't ordinarily show up very well.Â

Â
Question to the HOT folks -- is there a way to 
specify the date of DG imagery we access through 
the proxy server?, Some of the May 8 imagery is 
starting to come up over the May 3 imagery without me telling it to. Â


Best regards,

John
___ 
HOT mailing list HOT@openstreetmap.org 
https://lists.openstreetmap.org/listinfo/hot



___
HOT mailing list
HOT@openstreetmap.org
https://lists.openstreetmap.org/listinfo/hot

Re: [HOT] landslides and imagery

[Heather Leson]

HI Nama and John, while it is not satellite imagery, I've been following
the #nepalphotoproject. This collective seems to be documenting many
regions. With the quakemap and this photo stream could the landslide areas
be identified then verified with the imagery?

Thanks for all that you are doing

Heather
On May 10, 2015 4:52 AM, "Nama Budhathoki"  wrote:

> Hi John,
>
> Landslide is one of the major issues. We can try to use local knowledge
> for groundtruthing.
>
> Nama
>
>
>
> On Sun, May 10, 2015 at 6:34 AM, john o'l  wrote:
>
>> I've been focusing on landslides and have located several score that
>> appear recent.  Of these, a few are pre-quake and appear relatively stable,
>> some are pre-quake but appear reactivated and many appear to have been
>> associated with the quake and/or aftershocks. I've mapped several dozen of
>> these so far. In my next email, I'll cover why you won't find them in
>> osm... yet. For this one, I'd like to stick to post quake imagery and some
>> of its quirks.
>>
>> There is an inhabited hillside that had numerous landslides, some predate
>> the quake, but most are presumably related. So far I've mapped about half
>> of them, those that are largest or appear to threaten buildings and
>> pathways. There is Digital Globe imagery available from May 3 and May 8. It
>> looks like QGIS easily operates with more than one coordinate system at a
>> time. The center of a large boulder in the May 3 imagery (Longitude,
>> Latitude; WGS84 EPSG:3857 x,y) is at
>> 85.85659,27.83609;9557511.789,3228324.329, in the May 8 imagery it is at
>> 85.85669,27.83656;9557522.728,3228382.865. Mind you, this is not a
>> complaint, rather it is a concrete example of the variability with this
>> recent imagery.
>>
>> A more extreme example is a slide that appeared to be partially blocking
>> a stream in the May 3 imagery 85.90258,27.87818;9562631.312,3233623.303; --
>> it was completely obscured by a hillside in the May 8 imagery (probably
>> taken from a more northerly or northwesterly vantage point.)
>>
>> Downslope (westward) from a likely reactivated slide located at
>> 85.81987,27.90810;9553423.739,3237391.771  is a remote area that appears
>> very hard hit.  The May 8th imagery is mostly clouds, but the May 3rd
>> imagery shows a blue rooftop at 85.80644,27.90818;9551929.301,3237402.414,
>> it looks like there are several large boulders in the immediate area and
>> there is not much left to tell there were more than 20 buildings nearby.
>> While the boulders may have contributed, at the moment I think it is
>> probable that the shaking itself was mostly responsible for the extreme
>> level of destruction.
>>
>> One advantage of different acquisition angles is that some features may
>> be discernible on slopes that don't ordinarily show up very well.
>>
>> Question to the HOT folks -- is there a way to specify the date of DG
>> imagery we access through the proxy server?, Some of the May 8 imagery is
>> starting to come up over the May 3 imagery without me telling it to.
>>
>> Best regards,
>>
>> John
>>
>> ___
>> HOT mailing list
>> HOT@openstreetmap.org
>> https://lists.openstreetmap.org/listinfo/hot
>>
>>
>
>
> --
> 
> Nama R. Budhathoki, Ph.D.
> Executive Director, Kathmandu Living Labs *(www.kathmandulivinglabs.org
> )*
> Cell: 977-9803571739
> Office: 977-6205000
>
>
>
>
>
> ___
> HOT mailing list
> HOT@openstreetmap.org
> https://lists.openstreetmap.org/listinfo/hot
>
>
___
HOT mailing list
HOT@openstreetmap.org
https://lists.openstreetmap.org/listinfo/hot

Re: [HOT] landslides and imagery

[Nama Budhathoki]

Hi John,

Landslide is one of the major issues. We can try to use local knowledge for
groundtruthing.

Nama



On Sun, May 10, 2015 at 6:34 AM, john o'l  wrote:

> I've been focusing on landslides and have located several score that
> appear recent.  Of these, a few are pre-quake and appear relatively stable,
> some are pre-quake but appear reactivated and many appear to have been
> associated with the quake and/or aftershocks. I've mapped several dozen of
> these so far. In my next email, I'll cover why you won't find them in
> osm... yet. For this one, I'd like to stick to post quake imagery and some
> of its quirks.
>
> There is an inhabited hillside that had numerous landslides, some predate
> the quake, but most are presumably related. So far I've mapped about half
> of them, those that are largest or appear to threaten buildings and
> pathways. There is Digital Globe imagery available from May 3 and May 8. It
> looks like QGIS easily operates with more than one coordinate system at a
> time. The center of a large boulder in the May 3 imagery (Longitude,
> Latitude; WGS84 EPSG:3857 x,y) is at
> 85.85659,27.83609;9557511.789,3228324.329, in the May 8 imagery it is at
> 85.85669,27.83656;9557522.728,3228382.865. Mind you, this is not a
> complaint, rather it is a concrete example of the variability with this
> recent imagery.
>
> A more extreme example is a slide that appeared to be partially blocking a
> stream in the May 3 imagery 85.90258,27.87818;9562631.312,3233623.303; --
> it was completely obscured by a hillside in the May 8 imagery (probably
> taken from a more northerly or northwesterly vantage point.)
>
> Downslope (westward) from a likely reactivated slide located at
> 85.81987,27.90810;9553423.739,3237391.771  is a remote area that appears
> very hard hit.  The May 8th imagery is mostly clouds, but the May 3rd
> imagery shows a blue rooftop at 85.80644,27.90818;9551929.301,3237402.414,
> it looks like there are several large boulders in the immediate area and
> there is not much left to tell there were more than 20 buildings nearby.
> While the boulders may have contributed, at the moment I think it is
> probable that the shaking itself was mostly responsible for the extreme
> level of destruction.
>
> One advantage of different acquisition angles is that some features may be
> discernible on slopes that don't ordinarily show up very well.
>
> Question to the HOT folks -- is there a way to specify the date of DG
> imagery we access through the proxy server?, Some of the May 8 imagery is
> starting to come up over the May 3 imagery without me telling it to.
>
> Best regards,
>
> John
>
> ___
> HOT mailing list
> HOT@openstreetmap.org
> https://lists.openstreetmap.org/listinfo/hot
>
>


-- 

Nama R. Budhathoki, Ph.D.
Executive Director, Kathmandu Living Labs *(www.kathmandulivinglabs.org
)*
Cell: 977-9803571739
Office: 977-6205000
___
HOT mailing list
HOT@openstreetmap.org
https://lists.openstreetmap.org/listinfo/hot

[HOT] landslides and imagery

[john o'l]

I've been focusing on landslides and have located several score that appear
recent.  Of these, a few are pre-quake and appear relatively stable, some
are pre-quake but appear reactivated and many appear to have been
associated with the quake and/or aftershocks. I've mapped several dozen of
these so far. In my next email, I'll cover why you won't find them in
osm... yet. For this one, I'd like to stick to post quake imagery and some
of its quirks.

There is an inhabited hillside that had numerous landslides, some predate
the quake, but most are presumably related. So far I've mapped about half
of them, those that are largest or appear to threaten buildings and
pathways. There is Digital Globe imagery available from May 3 and May 8. It
looks like QGIS easily operates with more than one coordinate system at a
time. The center of a large boulder in the May 3 imagery (Longitude,
Latitude; WGS84 EPSG:3857 x,y) is at
85.85659,27.83609;9557511.789,3228324.329, in the May 8 imagery it is at
85.85669,27.83656;9557522.728,3228382.865. Mind you, this is not a
complaint, rather it is a concrete example of the variability with this
recent imagery.

A more extreme example is a slide that appeared to be partially blocking a
stream in the May 3 imagery 85.90258,27.87818;9562631.312,3233623.303; --
it was completely obscured by a hillside in the May 8 imagery (probably
taken from a more northerly or northwesterly vantage point.)

Downslope (westward) from a likely reactivated slide located at
85.81987,27.90810;9553423.739,3237391.771  is a remote area that appears
very hard hit.  The May 8th imagery is mostly clouds, but the May 3rd
imagery shows a blue rooftop at 85.80644,27.90818;9551929.301,3237402.414,
it looks like there are several large boulders in the immediate area and
there is not much left to tell there were more than 20 buildings nearby.
While the boulders may have contributed, at the moment I think it is
probable that the shaking itself was mostly responsible for the extreme
level of destruction.

One advantage of different acquisition angles is that some features may be
discernible on slopes that don't ordinarily show up very well.

Question to the HOT folks -- is there a way to specify the date of DG
imagery we access through the proxy server?, Some of the May 8 imagery is
starting to come up over the May 3 imagery without me telling it to.

Best regards,

John
___
HOT mailing list
HOT@openstreetmap.org
https://lists.openstreetmap.org/listinfo/hot