pseabury
Tourist
Reged: 04/14/05
Posts: 241
|
|
Update :
Quote:
Also, I've already used a couple of gigs of bandwidth this month....I have plenty to go around, but would rather folks update their links to the KMZ from the KML as the KML accounts for about 60% of my bandwidth. I've seen an issue where some versions of the client doesn't like the KMZ (no subfolders show up), but if your's works with KMZ please switch. Again, the data in both is identical, KMZ is just compressed.
Link to KMZ
<a href="" target="_blank">Link to KML Removed</a>
This link kind of goes along with some of my previous work in this thread . I've written an app that pulls the lastest forecast tracks and predicted strength for tropical named storms (Depression, Storm, Hurricane) from the National Hurricane Center, and then writes the kml that represents the data. This link isn't as dynamic as I'd like, as the NHC only puts out a new forecast every 6 hrs or so, but it's the best I could do so far. It should be pretty cool to look at when we have a few storms roaring. I've also not tested it exhaustively, so my app may break if it encounters data outside the format that I expect. I would guess that I may run into 1 or 2 of these as the format that they provide isn't optimal for parsing. I wish they'd just put out a comma-separated list, but I guess beggars can't be choosers. The app is designed to fail gracefully though, so should it fail to write new output for some reason, the last good data should be available until the problem gets fixed.
The only thing I ask is that you not hammer my server by setting it to update every 1 second or whatever for two reasons. First, the link itself is not from home (actually on a decent server), but I'd still rather not drive up my bandwidth usage. Second, the data will only update every 6 hrs or so, so the 10 minute default refresh should do just fine. Enjoy.
[EDIT] - Forgot the screenshot. Just as I was taking the screenshot, NHC added Franklin back into the data. Kinda cool that it worked as planned.
[EDIT] - Fixed the link to the network link in the original post. Sorry about that!
Paul
Edited by pseabury (08/27/06 07:18 PM)
|
troxy18
Tourist
Reged: 06/30/05
Posts: 137
|
|
Good use of information, looks pretty neat if you include a cloud overlay underneath or weather radar to make it look better. Like this 
|
pseabury
Tourist
Reged: 04/14/05
Posts: 241
|
|
Definitely Troxy. I think I could integrate boatloads of other tropical data along with the forecast tracks. I'm not sure I want to force that on everyone or bloat the kml with it though. Depends on the response I guess. If that's what people want, I can surely do it fairly easily. I guess the layers that a user doesn't want can be turned off by them eh? I think you just talked me into it...now lets see how much time I have after work this week....might have to wait until the weekend for the "Tropical Portfolio".
Paul
--------------------
There are 10 kinds of people in this world...
those that understand binary, and those that don't.
|
troxy18
Tourist
Reged: 06/30/05
Posts: 137
|
|
Or you could post a link to the original posting for the sat weather image I used that way people could see how different layers and overlays in google earth work together. Just think of the possibilities
|
3Dsolar
Tourist
Reged: 07/18/05
Posts: 43
Loc: UK & France
|
|
very good job !!
|
Mike_in_Florida
Storm Chaser
Reged: 07/25/05
Posts: 970
Loc: Florida, USA
|
|
Sounds great! Living in Florida this is very useful!
Thanks for working on it...
--------------------
|
pseabury
Tourist
Reged: 04/14/05
Posts: 241
|
|
No problem, I enjoy doing this sort of stuff. I'll likely keep improving it, but since it's a network link, that should be transparent to those using it.
--------------------
There are 10 kinds of people in this world...
those that understand binary, and those that don't.
|
SoFlaChris
Tourist
Reged: 07/04/05
Posts: 20
Loc: Boynton Beach, FL
|
|
Very slick... Thanks for taking the time to share this! Any chance that the model source name could be added to each run? (GFDL, BAMD. etc?)
TIA,
Chris
Edited by SoFlaChris (07/25/05 03:27 PM)
|
pseabury
Tourist
Reged: 04/14/05
Posts: 241
|
|
Currently that info IS in the data. You can expand "NHC Models" entry on your places menu and there will be a folder for each storm in the current report. You can then expand each storm's folder and see that each model has its own folder as well. Also, single clicking on a point in the track lines will bring up the info for that particular point, Including the model name. Single Clicking on the white point (the last real, not forecast, location) will bring up the raw Initial Conditions report for that storm.
Right now I'm currently working on adding a Sea Surface Temperature (SST) overlay from the raw satellite data. I'm taking the .25x.25 degree points from the GRIB (Gridded Binary) files, converting the raw data to a temperature C, and then creating an image from it. It's a lot of work, but I can control how it's drawn with a lot more granularity. I'll add this to the network feed once it's done and stable.
Regards,
Paul
Edited by pseabury (07/25/05 06:09 PM)
|
TheSHAD0W
Explorer
Reged: 07/05/05
Posts: 177
|
|
If you do decide to integrate an overlay, I've attached the one I'd recommend you use. Maybe you should produce two versions, one that includes an overlay and one that doesn't.
I'd also like to suggest (at least with this overlay) you increase the line width to make things easier to see.
|
pseabury
Tourist
Reged: 04/14/05
Posts: 241
|
|
Your attached overlay looks like an Infrared of the atlantic/caribbean. I intend to eventually include a plethora of data, all related to tropical activity, to this feed. I'm just doing it one step at a time. I'll look at the source of that overlay data and see if I can find a higher resolution one before I include it. Not that it loks bad, but I'm going to try to include the highest res stuff for all the data in the link. Thanks for your input...hopefully towards the end of this week there'll be LOTS more to this link.
[EDIT] - Also keep in mind that I mean highest resolution both spatially and temporally.
Paully
Edited by pseabury (07/25/05 06:15 PM)
|
SoFlaChris
Tourist
Reged: 07/04/05
Posts: 20
Loc: Boynton Beach, FL
|
|
DOH... I should have thought to open the folder! This is great. Keep it coming.
--------------------
chris :: wired
|
pseabury
Tourist
Reged: 04/14/05
Posts: 241
|
|
No Problem at all. If people don't typically expand their placemarks, then I should provide some other indication...maybe a legend like my other hurricane display has. That's the kind of stuff I'm trying to find out from people that post. Thanks Chris.
Paul
|
CapeVerde
Tourist
Reged: 07/13/05
Posts: 24
|
|
Most excellent! Some of us weatherbugs living on the coast will be using this layer a lot. Thanks!
|
pseabury
Tourist
Reged: 04/14/05
Posts: 241
|
|
Thanks CapeVerde, much appreciated.
I just updated the link slightly. Now the generated KML makes the placemarks and lines with an altitude, to prevent the lines from dissapearing into the ocean/land when the distance between points in very far. I didn't want to make the lines relative to ground in case they passed over a bump in elevation. It still may not be perfect, but it's at least better. I also made the lines 1 pixel thicker and scaled up the point graphics slightly.
BTW, I was talking to a meteorologist about the format of the data, as well as the storms named "Invest". I thought they were bogus, but he assured me that "Invest" is typically a name given to a storm in the preliminary stages of tracking it. I suppose they don't want to waste a name from the list on a storm of this nature, but then again, they want to give it a name. There may be more to it than that, but that's what I know about it.
Paul
|
LynnA
Tourist
Reged: 07/26/05
Posts: 5
Loc: Naples, FL
|
|
You guys (and GE) are amazing. I'm a newbie but a huge tropical weather enthusiast and this is so helpful.
I notice when I import the links in GE they end up in temporary places. Shouldn't they be overlays?
Thanks for all of your hard work.
Lynn
|
pseabury
Tourist
Reged: 04/14/05
Posts: 241
|
|
Lynn,
You can just Drag and Drop the Placemark(Folder, whatever) to your "My Places" tree. Then you can save it manually, or when GE exits it will ask you to save if you haven't already.
Paul
|
CapeVerde
Tourist
Reged: 07/13/05
Posts: 24
|
|
The Invest storms are far from bogus. They are tropical waves that appear to have the potential to develop into tropical depressions, storms, or hurricanes.
Avid weatherwatchers follow them closely, as does the the National Hurricane Center. An Invest is something they spend time and manpower investigating with flights to gather atmospheric conditions.
A new Invest, 93L, was identified last night, and the huge wave coming off Africa right now is almost certain to become Invest 94L within a day or two.
They don't actually give a storm a name until it reaches tropical storm status. It has to be identified as a tropical depression before that and just given a different number. If Invest 92L becomes a tropical depression, it will be called TD8.
Being able to track these Invests on Earth Google thanks to your work is just awesome, in my opinion.
CV
|
pseabury
Tourist
Reged: 04/14/05
Posts: 241
|
|
Thanks CV......it was pretty cool to get up this morning and see two new "Invests" steaming westward in the tropical atlantic. I'm still working on the raw SST, but be patient...I'll get it soon enough.
Paul
|
CapeVerde
Tourist
Reged: 07/13/05
Posts: 24
|
|
In addition to your layer, I've downloaded a layer for global satellite infrared radar which depicts cloud temperatures (colder meaning bigger and more stormy). I got it here:Global Infrared Satellite Images
The combination of your layer with this makes an excellent way of tracking storms and their intensity.
When you develop your SST layer, it's going to be an incredibly valuable tool in predicting increases in storm intensity.
CV
|
pseabury
Tourist
Reged: 04/14/05
Posts: 241
|
|
That is a nice IR layer. I changed some of the overlays to auto update, but other than that kept it as is. Looks good, and thanks.
--------------------
There are 10 kinds of people in this world...
those that understand binary, and those that don't.
|
pseabury
Tourist
Reged: 04/14/05
Posts: 241
|
|
First cut at SST is done. I haven't added it to the link yet because I've got to run for a couple hours and want to debug/polish it up a bit. Parsing those gribbed RAW data files isn't difficult per se, but it does make you want to kill byte arrays.  Look for it to automatically be added either tonight ir tomorrow morning when your link refreshes. Here's a screenshot of the Alpha...
--------------------
There are 10 kinds of people in this world...
those that understand binary, and those that don't.
|
LynnA
Tourist
Reged: 07/26/05
Posts: 5
Loc: Naples, FL
|
|
Thanks. Appreciate the tip. I'm starting to get the hang of it. Unfortunately, I am now having fetching errors when the NHC tracks update. I'll figure it out yet with the help of you great people!
Again, thanks,
Lynn
|
pseabury
Tourist
Reged: 04/14/05
Posts: 241
|
|
Lynn,
You may be getting fetch errors because I'm working on it. I'm sorry about that. If the kml is bad on a refresh, it should be fixed by the next timed update (10 mins I think). At least tonight, and probably at certain times tmw, you may get this same error. Hopefully it will be worth the fetch errors once I quit editing it.
|
LynnA
Tourist
Reged: 07/26/05
Posts: 5
Loc: Naples, FL
|
|
Thanks for letting me know. I thought it was my computer or something I was doing wrong. Keep up the great work!!!
Lynn
PS Is the Invest data missing because you are working on it as well?
Edited by LynnA (07/29/05 08:07 PM)
|
pseabury
Tourist
Reged: 04/14/05
Posts: 241
|
|
The track models (including Invest) aren't put out by the NHC in the most reliable format. For instance.....the main reports that they put out are at 00Z, 06Z, 12Z, and 18Z. Four times per day, 6 hrs apart. Unfortunately, right now I only check for the latest one of those 4 reports. They also put out intermediate reports at say 01Z, 07Z, maybe 11Z, etc. . I'm not sure why they do this, other than the models for a particular storm may not be ready yet. In any case, the way I coded by client was the check for the very latest file by timestamp (the youngest file), the check to see if it one of the {00,06,12,18) set. If it isn't, then I check the next oldest file until I find one that falls into the (00,06,12,18) set so that I have the "official" report. The problem is that sometimes the official 6hr ones don't get updated with the proper information, they just put out one of those intermediate files. So to maybe make it a little more clear (I tend to ramble) here's a typical situation if say we are tracking 3 storms(2 invests, and 1 proper named):
00Z - Invest1, Invest2, Franklin
06Z - Invest2, Franklin
07Z - invest1
12Z - Invest1, Invest2, Franklin
18Z - Invest1, Invest2, Franklin
00Z - Invest1, Franklin
01Z - Invest2
So you can see that sometimes all of the storms aren't included in a single 6hr report. I will be working on this problem tomorrow. Odds are that if you're missing a storm that you think should be there, it will be there upon the next update.
On another note, I've upped the resolution of the SST overlays DRAMATICALLY and they look great now.....I'm going to try and get the Tropical Atlantic built into the feed tonight.
Regards,
Paul
|
pseabury
Tourist
Reged: 04/14/05
Posts: 241
|
|
Ok, the first cut at the SST data is now in the Network Link. If you just let the link refresh, the SST data should show up. It's very late, so there may be bugs, but I'm sure I'll hear from you if there are. You can see the improvement in resolution in the screenshot below when compared to the screenshot above. I did have some problems with GE complaining about the image format, but that just magically dissapeared....at least for me. Let me know.
--------------------
There are 10 kinds of people in this world...
those that understand binary, and those that don't.
|
LynnA
Tourist
Reged: 07/26/05
Posts: 5
Loc: Naples, FL
|
|
Paul:
Thanks so much. I know there are quite a few of us at Storm2K (weather board) who are using this product and we really appreciate all the work you are puttiing into it! Great job.
Lynn
|
pseabury
Tourist
Reged: 04/14/05
Posts: 241
|
|
No problem at all. I enjoy solving interesting problems. I'm looking for fairly high-res Visible, Infrared, and Water Vapor imagery now to add to this dataset. It'd be nice if it was geocoded or GRIBBed so that it fits nice and snug without hand fitting it. If anyone has any suggestions please let me know. Also, I do have the capability to easily add more sections of SST if anyone wants it.....I have the raw data for the entire globe, and I wrote a routine to generate bitmaps on the fly from input coordinates....so it's rather easy. Just let me know.
Paul
|
rodziolkow
Tourist
Reged: 02/04/05
Posts: 4
|
|
Very impressive! It is cool to see how you and pseabury collaborated to make layers that combine to present a very powerful image. Great work!
|
Bamastormchaser
Tourist
Reged: 07/30/05
Posts: 9
|
|
Very nice. Is there any way you can add more data feeds into the selection? Just curious.
--------------------
Joseph B. Adair
www.hardcoreweather.com
Mississippi State Meteorology Student
WBRC-FOX 6 Storm Chaser
|
pseabury
Tourist
Reged: 04/14/05
Posts: 241
|
|
Absolutely, just let me know what you are looking for and I'll definitely look into integrating it. I was looking at wind shear and vorticity overlays yesterday, but wasn't sure how many people may be interested in that. I don't want to add unneccessary things that just clutter up the display and network link, but I DO want things that are helpful in providing one-stop shopping for the tropical watchdogs.
Paul
|
Bamastormchaser
Tourist
Reged: 07/30/05
Posts: 9
|
|
Thanks Paul! I am interested in the wind shear/wind shear tendancy maps. You can take a look here at alot of cool stuff to integrate....
http://www.hurricanehollow.org/
Thanks! 
--------------------
Joseph B. Adair
www.hardcoreweather.com
Mississippi State Meteorology Student
WBRC-FOX 6 Storm Chaser
|
pseabury
Tourist
Reged: 04/14/05
Posts: 241
|
|
Yes, those are the exact images I was testing out yesterday from http://cimss.ssec.wisc.edu . Not the best resolution, but as long as they aren't floaters, then we can use them. I'd also really like to lose the lat/lon lines in the images to make it look cleaner, but that's all I've found so far. I've got some stuff to do today outside the house, so I may not get to it until tonight or tomorrow, but keep an eye out for the next interation of the link.
Also, sorry if some of you had problems with the link and/or images last night through this morning. I was running the debug version of my program that builds the images and kml which wasn't the latest build. I had only rebuilt the release version with the latest changes. Anyway, it should be popping out the latest stuff now. I'll keep you posted.
|
Bamastormchaser
Tourist
Reged: 07/30/05
Posts: 9
|
|
Previous
Index
Next
Threaded
View in Google Earth
Edit
Reply
Quote
