global clouds WMS?

[John Maurer]

Hi IOOS Techies,
Anybody know of a global clouds composite that is provided via WMS, supports spherical mercator (EPSG:3857), and is regularly updated? This is difficult to come by, though I've had several close encounters over the years. It would be a great way for our users throughout the insular Pacific to visualize tropical cyclones via our map portal (Voyager), such as recent Typhoon Vongfong that crossed over Guam on its way to Japan.

The ones I've tried:

1. NASA SIVO provides a good WMS-T (layer: "cpc") that is updated every half hour, but only supports a lat/lon projection (EPSG:4326) and does not have staff to update their MapServer config. I've tried re-serving and re-projecting this via our local MapServer installation, but the additional throughput renders overlays painfully slow.
2. NOAA nowCOAST provides one (GetCap, layers: "RAS_GOES" [visible], "RAS_GOES_I4" [infrared]), but it only covers the 50 states, so nothing west or south of Hawaii.
3. Iowa State provides GOES WMS services but only for regional subsets: no global composite.
4. OpenWeatherMap has a global clouds WMS (GetCap, layers: "clouds", "clouds_cls" [classic]), but it's low resolution (and kinda strange looking).
5. Similarly, Pacific Disaster Center (PDC) provides a low resolution global composite (GetCap, layer: "0"), but it is rather coarse and updates only every 3 hours.
6. Google Maps API seemed to provide the silver bullet a couple years ago, but their cloud overlay cannot be reliably refreshed and they are discontinuing support for this layer next June.
   

Just wondering if any of you have other recommendations?
Cheers,
John Maurer, PacIOOS

[Charlton Galvarino]

Hi, John.  Coincidentally I stumbled across this today, http://tropic.ssec.wisc.edu/.  The KML link is an almost overwhelming catalog of real time datasets.  Some may or may not have WMS underneath, but I did see several cloud cover layers.  Maybe there is a bone or two in there to aid you in your quest.

 

--
You received this message because you are subscribed to the Google Groups "ioos_tech" group.
To unsubscribe from this group and stop receiving emails from it, send an email to ioos_tech+...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Charlton Galvarino]

If the KML doesn’t have what you need, you might be able to take care of your NASA SIVO projection limitations by putting http://mapproxy.org/ between your client and the WMS-T.  I have had *amazing* success w/ mapproxy.  It even allows my OpenLayers app to hit a Google tile cache (EPSG:3857) as if it were hitting an EPSG:4326 WMS dataset.  I.e. OpenLayers is making an EPSG:4326 GetMap request to a Google tile cache EPSG:3857 thanks to a MapProxy monkey in the middle.

[Jeff de La Beaujardiere - NOAA]

John,

I am the person who established the NASA WMS you mentioned, though I believe it's only a plain WMS rather than tiled. This would have been in 2005 or so, before I left NASA for NOAA. However, I am still friends with the group, and *might* be able to convince the local sysadmin to make the config change to support the additional projection if you provide some specific instructions (i.e., open this file, add this line here, etc.) I have not tinkered with mapserver code & config since 2007 so I don't remember (and they may still be running the old version). 

Regards,

Jeff DLB

Jeff de La Beaujardiere, PhD 

NOAA Data Management Architect and 
Acting Director, Technology, Planning & Integration for Observations 
1335 East-West Hwy, Silver Spring MD 20910 USA 
+1 301 713 7175 (NESDIS/OSD/TPIO - SSMC1/5236) 

ORCID: http://orcid.org/0000-0002-1001-9210

--

[Rich Signell]

There is this modis terra tile service, here accessed/visualized using
the British Met Office Cartopy package in Python:
http://nbviewer.ipython.org/gist/rsignell-usgs/c1b8082eabb934e63fb1

--
Rich Signell
81 Queen St
Falmouth, MA 02540

[Dan Ramage]

John,

Another option, if you have map server running on your end, is to use your map server to do the re-projecting. I've done that with some WMS feeds that didn't have spherical mercator.

[John Maurer]

Thanks all for your feedback. Replies below...

Jeff DLB: Wow, cool! I will e-mail you separately to explain how MapServer can be configured to support EPSG:3857. This might be the best solution.

Charlton: I tried http://tropic.ssec.wisc.edu/real-time/kml/CIMSS-TCmain.kml. I could not find a global composite, only regional subsets. And even these contain coastlines, labels, and other undesirable map information in addition to just the cloud imagery. Thanks for pointing out MapProxy. This sounds promising in terms of speeding up the process of re-serving and re-projecting the NASA SIVO WMS, which I will investigate further if Jeff DLB hits any snags.

Pete Giencke (Google): Recommended http://mw1.google.com/mw-weather/clouds/root.kmz as possible workaround. This is what powers the clouds overlay in Google Earth. He said it might be finicky in Google Maps, and I can confirm that some of the super tiles fail to load without panning/zooming around. Also, this is fairly low resolution imagery compared to NASA SIVO WMS.

Rich: Thanks for the Python tip. However, I'd prefer to use a cloud-specific overlay rather than MODIS imagery that includes land and ocean. That would essentially overlay a basemap on top of my existing basemap and mask anything in between (e.g. SST, waves, whatever).

Dan: As I mentioned, I'm already using MapServer to do the re-projecting on my end as you suggest, but it is painfully slow to render tiles using this method. Charlton's MapProxy idea could help speed this up...

Cheers,

John

[Dan Ramage]

John,

I guess I just read right through that. The Map Proxy package is something I am interested in as well. Currently we use TileCache here to relieve the server load for map server, however it is 

Not without it's own issues.

Dan

[John Maurer]

Hi Jeff,
That's great news. To support additional projections, you would simply need to update the MapServer config file at http://map.nasa.gov/data/cpc/cpc.map. Within the WEB > METADATA section near the top you add this line:

"wms_srs" "EPSG:4326 EPSG:3857"

There is a catch, however. If the PROJ.4 "epsg" file that MapServer uses for pre-defined projection definitions is old, it may not contain EPSG:3857 yet and WMS will fail with an error. In this case, all you need to do is find and update the "epsg" file that MapServer is using with the following:

# Google Mercator projection
<3857> +proj=merc +a=6378137 +b=6378137 +lat_ts=0.0 +lon_0=0.0 +x_0=0.0 +y_0=0 +k=1.0 +units=m +nadgrids=@null +no_defs  <>

Depending on how PROJ.4 was installed, here are a couple suggestions of where to look for this "epsg" file:

/usr/local/share/proj/epsg
/usr/local/FWTools/share/proj/epsg

Hope you have success! Let me know if I can be of further assistance. I've forwarded below the last message I received Monday from NASA SIVO on this request.
Cheers,
John

---------- Forwarded message ----------

From: Mitchell, Horace G. (GSFC-6064) <horace.g...@nasa.gov>
Date: Mon, Oct 13, 2014 at 4:28 PM
Subject: Fwd: Global Clouds WMS: EPSG:3857 projection support?
To: John Maurer <jma...@hawaii.edu>
Cc: "Considine, David B. (HQ-DK000)" <david.b....@nasa.gov>, "Robert....@jpl.nasa.gov" <Robert....@jpl.nasa.gov>, "Lara.C...@gsfc.nasa.gov" <Lara.C...@gsfc.nasa.gov>, "Clune, Thomas L. (GSFC-6060)" <thomas....@nasa.gov>

John,

I’m afraid I don’t have good news for you on this issue.  This server was developed years ago and is not currently maintained by anyone.  The fact that it still works at all is amazing.  The original developer does not work here or in this field anymore, so there’s no chance of an update.

My initial reaction to your request was that the implementation of EPSG:3857 would be a lot of work, since new code for that projection would have to be implemented in our server.  Upon closer reading, I realized that you thought that the data was being served by MapServer, so that a configuration file change was all that was needed.  This may be true, but I cannot verify it.  I think that the developer did use MapServer for some datasets, but not all, and implementing EPSG:3857 throughout the server is not within the scope of our current work.

Sorry I can’t help.

Horace

On Oct 10, 2014, at 2:57 PM, John Maurer <jma...@hawaii.edu> wrote:

Dear map.nasa.gov,

Sorry, I wasn't sure who to e-mail specifically. I tried Tom Clune earlier this week but have not heard back. Please see my message below hoping you could add an additional map projection (EPSG:3857) to your global clouds Web Map Service (WMS). If so, we'd like to incorporate this layer in our integrated map viewer here at the Pacific Islands Ocean Observing System (PacIOOS) (http://pacioos.org/voyager/). Thanks for your help!
Cheers,
John Maurer
Data System Engineer
Pacific Islands Ocean Observing System (PacIOOS)
University of Hawaii at Manoa

---------- Forwarded message ----------
From: John Maurer <jma...@hawaii.edu>
Date: Mon, Oct 6, 2014 at 3:08 PM
Subject: Global Clouds WMS: EPSG:3857 projection support?
To: Thomas....@nasa.gov

Dear Thomas Clune,

I recently found and am hoping to integrate the following global clouds WMS from your lab (NASA GSFC SIVO):

http://geocommons.com/overlays/148081

http://wms.gsfc.nasa.gov/cgi-bin/cpc-wms.cgi?service=WMS&request=GetCapabilities&version=1.1.1

http://map.nasa.gov/data/cpc/cpc.map

This is the first WMS providing global cloud coverage that I have ever found and was happy to see this! I'm hoping to incorporate it into our integrated map viewer called Voyager at the Pacific Islands Ocean Observing System (PacIOOS) of the University of Hawaii: http://pacioos.org/voyager/. Currently, I only provide cloud coverage from NOAA nowCOAST WMS, which does not extend west of Hawaii. We support other regions in the Pacific (CNMI, FSM, Guam, American Samoa, Palau, etc.), which is why global cloud coverage would be a big plus for us. For instance, Voyager could not suitably demonstrate Typhoon Vongfong passing over Guam this weekend.

My question is this. Could you please add support for EPSG:3857 (Google Spherical Mercator) in your WMS? Currently, you only advertise EPSG:4326 (latlong) in your GetCapabilities document. To do this, you would simply need to update the MapServer config file at http://map.nasa.gov/data/cpc/cpc.map. Within the WEB > METADATA section near the top you could add:

ows_srs "EPSG:4326 EPSG:3857"

Otherwise, I cannot import the WMS layer into our Google Maps based visualization system. All GetMap requests with SRS=EPSG:3857 fail with the following error:

<ServiceException code="InvalidSRS">
msWMSLoadGetMapParams(): WMS server error. Invalid SRS given : SRS must be valid for all requested layers.
</ServiceException>

Thanks for your help on this!
Cheers,
John Maurer
Data System Engineer

Pacific Islands Ocean Observing System (PacIOOS)
University of Hawaii at Manoa