Updated information RSS of this wikiHistory of "Frontpage" is here.
Last Update (diff) of "Frontpage"
Added Line is this color.
Deleted Line is this color.
---- Keep this link on top ***[[TCS08_Project>http://patka.nmt.edu]] ---- -&color(blue){3/27/09 MDT 13:58- I finished tracking vorticity spots for EPAC between April and November. I don't believe there are any more significant vorticity events which have not be labeled and tracked. Next week, I will glance over the WPAC spots and see if there are any I have missed. I have been thinking: would it be useful to post the trackfiles for the storms online? When I track storms, I often use other people's coordinates as a comparison to make sure that my tracks are accurate. Do you think that someone else might benefit or appreciate seeing our coordinates? Anyway, that is all I have to say. Have a good weekend everyone. -Anna} +&color(blue){3/27/09 MDT 13:58- I finished tracking vorticity spots for EPAC between April and November. I don't believe there are any more significant vorticity events which have not be labeled and tracked. Next week, I will glance over the WPAC spots and see if there are any I have missed. I have been thinking: would it be useful to post the trackfiles for the storms online? When I track storms, I often use other people's coordinates as a comparison to make sure that my tracks are accurate. Do you think that someone else might benefit or appreciate seeing our coordinates? Let me know what you think and how feasible this is. -Anna} ---- &color(green){3/16/09 MDT 8:24- Thanks, Anna! The scripts to generate labelled plots should be working. I made plots for all the stuff I did earlier, and you might want to regenerate them for the storms you touched up. Also, the timeplot stuff should be working. -Mike} ---- &color(blue){2/6/2009 MDT 10:09- I touched up the tracks that you did earlier this week, Mike. Overall, they were pretty good. I made some changes to &bold(){11lkyle}, &bold(){13emarie} and &bold(){ 15enorbert}. I haven't gotten around to tracking spots yet, but I will work on that after the break. In order to do this, I will need labeled survey plots. Do we have these reliabley yet? Thanks! -Anna} ---- &color(green){3/04/09 MDT 8:49- All named EPAC storms have been tracked (though, not very carefully, due to time constraints!) Also, timeplots.sh is working with EPAC, now (a little buggy, but ok for now). The EPAC storm tracks could use a close inspection. Also, the little spots, if there are any worth tracking, should be tracked. -Mike} ---- &color(blue){2/13/09 MDT 16:00- Today I continued tracking disturbances in the East Pacific. There is complete tracks and data now on 05ledouard, 09ehernan, 10eiselle, 06lfay, 11ejulio, 07lgustav, 08lhanna, 12ekarina and 09like. The EPAC page looks a little less pitiful now! This only brings me into September, but I expect that I will be finished tracking the EPAC in a few weeks. I still hope to work on making timeplots for EPAC, but I need to do some more evaluation before I go changing programs that other processes may require. After this is done, I'll need a new project! Does WPAC need re-tracking? -Anna} ---- &color(green){2/02/09 MDT 9:46- Ok Anna, typhoon now makes a file called storm_track.bak in the storm's directory before it opens the trackfile. I haven't fixed the START and END bugs, yet, but if you have trouble again, you can fall back on the backup. -Mike} ---- &color(green){2/02/09 MDT 8:33- Anna, sorry about that nasty bug in typhoon! I will try to look at it this week. I guess an initial fix would be to make backups of the trackfiles in the storm directory when typhoon starts up. I'll call them: storm_track.bak. It looks like you made quite a bit of progress on friday! I see you even made the labeled plots. The syntax for the program to generate more domain plots is:} tcssurvey_all.sh epac fnl 08 06 &color(green){I'm generating them up through November, now. Getting the month-summary program to work would be great! Then, the website would be mostly in working order. When you finish with the named storms in EPAC, you could look for long-lived spots and try to track those. -Mike} ---- &color(blue){1/30/2009 MDT 15:22- Hi! Today I tracked all of the storms in the East Pacific up until the beginning of August 2008 where the images ran out. Let me know what program generates our survey plots and I will run it myself from now on whenever I run out of data. The storms tracked include 04edouglas, 05enoname, 02lbertha, 06eelida, 07efausto, 03lcristobal, 08egenevieve and 04ldolly. I did not track 03ecristina because it was entirely out of our domain of interest. I had some troubles with Typhoon. In particular, the START and END buttons were giving me problems. Often, when I pushed START, all or most of my data would inexplicably be cropped (even data later than the date I was currently on.) Once, instead of cropping the data at all, it simply added more "#month date lat lon..." lines at the beginning of the trackfile. In each instance, it either destroyed my track or made doall.sh fail. The END button was also buggy, but less often and less disasterously. Occaisonally when I pressed it, my data would suddenly jump several months back. This was annoying, but easily worked with. The START button, however, was not worth messing with, so I started manually cropping data from the trackfile. I have no clue what might be the problem since everything seemed to work smoothly last week. I hope it is not too much trouble to work this issues out. I will be back in by Friday to continue storm tracking in the EPAC. I also hope to alter my month-summary program to work on the EPAC as well as WPAC. If there is anything else you would like me to work on, just let me know! See you next week! -Anna} ---- &color(green){1/26/09 MDT 12:24- Anna, I fixed some bugs in the website. webplot_update.sh is still a little shaky, but it is almost working. From now on, I'll be working on the West Pacific stuff, and will try to minimize the programming work. However, lemme know if more bugs turn up, and I'll try and fix 'em. -Mike} ---- &color(blue){1/23/08 MDT 4:10- After finally returning from last semester, I reoriented myself with the altered site by re-tracking &bold(){01ealma} and &bold(){01larthur} and I started the track of &bold(){02eboris}. Unfortunately I couldn't finish its track because we don't have any images yet for July 2008. I have fogotten what program generates the unlabeled images, so I will leave the website as is for now. I will definately be working on Fridays and I hope to find other times when I could drop in and do some tracking/debugging. Meanwhile, its good to be back! Lets hope for a good semester! -Anna} ---- &color(green){11/07/08 MDT 8:37- Well, there is still some activity in the west pacific. For instance, there is currently a named storm: &bold(24wtwentyfour). However, when I run todaysirjpeg.sh, weird things are happening. So, there are weird things at the top of our website! One thing that comes to mind is that we should be careful to keep as much of a storm's history as possible when tracking. That will allow us to get the most out of numerical analysis, since we will have lots of data for each storm. The early stages of named storm's lifetimes are important since they have not yet spun up, and likewise, the entire lifetimes of weak easterly waves are important because they never spin up. Finally, the dying phase of powerful storms are important, so we can see what processes are weakening them. I think we should keep the full tracks for named storms, but not their tcs or invest incarnations (although, make a note on the site of the previous incarnations of the storm, i.e. its tcs number). However, if a tcs never makes it to invest or named storm, keep the highest level incarnation (tcs or invest). Finally, if there is no tcs or invest, yet we can identify an easterly wave, lets use Anna's naming convention for easterly waves, shown in the right-hand column of the website. -Mike} ---- &color(green){10/30/08 MDT 10:30- Anna, I added the [START] button feature. It will truncate the trackfile immediately before the current time and update the trackfile box accordingly. If you make a mistake, just close and reopen the file. If you didn't save the error, the trackfile is still ok. Lemme know if you see bugs! -Mike} ---- &color(green){10/28/08 MDT 16:38- I made minor changs to doall.sh. Now it makes a textfile containing all the timeplot variables. This file is then sent to /home.gryphon/gfs/karen/ for storage. If there is any weirdness with doall, lemme know. Anna, I'll add that feature to typhoon ASAP! That's a great idea. I dunno how many times I've wanted to truncate the start of a file. Also, I'm glad you altered blobtrack_all like that. I bet it runs pretty quick, now. -Mike} ---- &color(blue){10/28/08 MDT 11:47- I finally checked out the new typhoon program. Its looking great, Mike! I hadn't even thought of the key-board shortcuts to jump to the beginning and end of a storm-track, but I find myself using it a lot! I did some minor tweaking on &bold(){21whigos} and &bold(){93winvest06} and some major tweaking on &bold(){94winvest05}. I also altered blobtrack_all.sh so that, if you run it with only a month and year argument, it will only run on FNL-data. If you add a "1" on the end, it will run for GFS-30 and 54. Since we aren't using GFS data at the moment, this should cut run-time enormously and hopefully make the program more convenient. Oh, and on that note, there is one more thing that would be nice to see in typhoon: we have an "end" button to end the storm and crop all data beyond. Would it be possible to have a "start" button to crop all data earlier? This would make it much easier to remove bad track data at the beginning of a storm. Let me know if it is feasible! Thanks! -Anna} ---- &color(green){10/24/08 MDT 16:34- I made minor changs to: tsctcanal.sh, tcstrack.sh, tcsplot.sh, tcsplot.template, tcscircdivplot.template, in order to standardize units in similar variables, and to fix the formula for absolute vorticity, which needed to be converted from deg2rad. This little change caused the spinup variables to perk up considerably and I reset the limits on the plots to accomodate this. The spinup variable now indicates spinup generally when it should. -Mike} ---- &color(green){10/21/08 MDT 17:19- Well, our new spot-tracker has arrived! Type 'typhoon' on the command line and see what comes up. You must be in a storm directory to start the program, but you can access all other storms (for the relevant file-type) from within the program. Also, new storms can be created. Thanks for all the good suggestions, Anna! -Mike} ---- &color(green){10/8/08 UTC 14:40- Touched up a few things today. No bugs! Looks like the typhoon season is coming ot a close rapidly. There are two new tcs's today, but they turned out to be copies of old ones! Nothing more. -Mike} ---- &color(green){10/7/08 UTC 15:00- I have started a backup of this wiki in /data.pt3/gfsteam_logs/wikilog. I will try to remember to add to it. Things seem to be working fine today. All tcs's have fizzled out, with the possible exception of &bold(){tcs054}, and its related &bold(){94winvest05}. The navy is only tracking those, for now. NOTE: I edited tcssatplot.sh to deleted all files in the /home.gryphon/gfs/.lyx/cache directory. This is to control the size of the gfs's gryphon directory, which had grown to 42G! -Mike} ---- &color(green){10/6/08 UTC 18:00- It looks like convert stopped functioning on the 4th of this month. None of the survey plots have been converted, and I see errors in the log for all other convert statements. Running doall on tcs49 to see what I get. Well, doall works ok, but spot_touch won't run, since none of the recent jpgs exist for the data type. I ran tcssurvey_all.sh for the month of October, and that allows spot_touch to run. Now, things seem ok. (However, cron might have trouble, still). There were a couple of missing data files today. They should have come in on 10/02 12h. I'm not sure what happened, but they are in, now. Other weirdness includes the fact that some of the gfs track files were missing datapoints right in the middle of the file! Hmm.... Well, things seem ok, now. -Mike} ---- &color(blue){10/3/08 UTC 21:00- I have ended &bold(){19wjangmi} and &bold(){21whigos} and purged a couple of worthless invests and tcs's today. Tracks for all of the other storms have been touched up. Keep an eye on &bold(){93winvest06}. It seemed to show up out of nowhere and has already developed a pretty strong vorticity center. We'll see if anything comes of it. In other news, WELCOME BACK GUAM TEAM! -Anna} ---- &color(green){10/1/08 UTC 17:40- There are no flights today. Anna, I took your cue and ended &bold(){19wjangmi} gfs 54. We should probably end the other two, as well. I purged &bold(){tcs060} and &bold(){90winvest}. Also, I ended &bold(){tcs048}, since it is being tracked as &bold(){21whigos}. I fixed another bug (one I introduced yesterday!). -Mike} ---- &color(blue){9/30/08 UTC 19:30- One more thing. I think it would be a good idea to keep a backup copy of this log somewhere safe to be certain that we don't lose all the observations and information we store here. As such, I am creating a file named &bold(){tcs_log.txt} in our www directory. I have copied everything below this entry into it. I will try to remember to copy new posts and entries into it at least once a week! -Anna} ---- &color(blue){9/30/08 UTC 19:15- So, it turns out that &bold(){tcs048} DID spin up to become the current &bold(){21whigos}. Of course, this storm's behavior is still far from normal. It displays considerably less circulation than most named storms and most of the circulation is still concentrated at lower levels. I am still curious if this has anything to do with the recent passage of &bold(){19wjangmi} through the area and if departing from &bold(){19wjangmi}'s track (as it is now doing) will boost its power.Unfortunately, &bold(){21whigos} is dead on track to leave our domain in the next couple of days, so I may never know. In other news, I attempted to alter and distinguish the tracks on &bold(){tcs050} and &bold(){tcs060}, but I am clueless. We will just have to wait and see what the Navy thinks. Until then, your guess is as good as mine. Additionally, I touched up &bold(){19wjangmi}'s gfs-tracks, but I am dubious that the stream of vorticity that it morphs into at the northern boundary of our domain can be considered a valid, distinct storm. I will wait to see what the navy says, but keep in mind that these tracks may need to be truncated once the fnl-track catches up a little more. Thats all for now! See you on Friday! -Anna} ---- &color(green){9/30/08 UTC 16:40- &bold(){19wjangmi} is of interest today. Well, I think I fixed the &bold(){utrans-blowup} bug today. It was an old bug, and dates back to June. The bug is actually in cdftable, which converts a text file (in this case, our handmade storm_track.txt files) into a candis file. The bug comes in when there is an empty line at the end of the text file. When cdftable finds one of these, it simply copies the data from the previous entry into a bogus entry where the extra space was. This causes the translations utrans, vtrans to blow up. The bug is still in cdftable (that is a job for another day), but I added a trackfile fixer line into tcstrack.sh that edits the trackfile and removes any empty lines. -Mike} ---- &color(green){9/29/08 UTC 15:30- &bold(){19wjangmi} is of interest today. I got the &bold(){utrans-blowup} on &bold(){tcs049}, gfs 54 and &bold(){21wtwentyone}, fnl today. I was running 3 instances of doall.sh and when I quit the process, I noticed that there were about 10 instances of tcstcanal running at once! I wonder if that might be part of the problem, since tcstcanal is what generates the .combo files. Well, I quit and waited for the other doall process to finish. Then, I deleted *track.cdf in the storm dir., but that didn't fix it. In fact, I had only one doall process running, and nothing else, and they still wouldn't work. I noticed that on &bold(){tcs049}, tcssatplot.sh and combine.sh were hung up, and doall wouldn't finish. I tried commenting out the combine.sh line in tcssatplot.sh, but got the same results. -Mike} ---- &color(blue){9/26/08 UTC 21:10- Today I simply touched up the storm-tracks of all active storms. In doing so, I encountered a fairly common error which will hereafter be referred to as the &bold(){utrans-blowup}. Although I have not yet found the bug causing this error, I believe that it involves the sharing of a filename between doall.sh and runblobs.sh causing an interference when two doall's or doall and runblobs are run simultaneously. For now, the solution to the error is to &bold(){delete the track.cdf file for the storm before re-running doall.sh}. Hopefully this information will expedite the debugging process as well as the storm-tracking process! Other than that, there is not much to report. &bold(){tcs048} does seem to be showing possible signs of spinup, but there are several unusual aspects to its behavior. For example, the circulation that is growing is currently very low-level and the gfs tracks show that it will stay there. Additionally, I wonder whether it is healthy for a storm to follow immediately behind another strong storm system. Might following behind &bold(){19wjangmi} hurt &bold(){tcs048}'s development??? Mike, you might be better qualified to answer that question for me! -Anna} ---- &color(green){9/26/08 UTC 14:30- &bold(){19wjangmi} is of interest today. I ended &bold(){98winvest05}, since it is at the western edge of the domain, and heading west. It has a dummy track of three points long in each model to prevent doall.sh from messing with it any further. &bold(){tcs049} is another little easterly coming into the domain. Not much of interest there, at present. &bold(){19wjangmi} is becoming quite impressive, with optimal conditions in moisture and rel. wind, though gfs 54 predicts a lessening in circulation before it reaches Taiwan. The gfs 54 prediction for &bold(){tcs048} indicates that the SST is having little effect on the storm. It is too early to tell, though. The SST peak is still 18 hours off. -Mike} ---- &color(green){9/25/08 UTC 15:05- &bold(){19wjangmi} is of interest today. Nice prediction work, Anna! It looks like &bold(){96winvest04} has become quite a storm. &bold(){97winvest04} is not being tracked by the navy today, so I think they must have dropped it. However, the related &bold(){tcs048} is still active and seems to exhibit favorable conditions for future spinup. &bold(){19wjangmi} is getting stronger and is headed for Taiwan. The gfs model predicts a peak in SST this afternoon at 18:00 UTC. Meanwhile, its moisture content is high and predictions show relative winds to stay below threshold for the next 48 hours. &bold(){tcs048} looks similar to &bold(){19wjangmi} did at this stage on jday 264. Remarkably, on that day, these two storms are in the exact smae location lat/lon, and they have nearly identical circulation, SST, moisture content, and relative winds. Only big differences are the entrainment in tcs048 is higher, and it is projected to go closer to Guam. Will we have another typhoon in 4 days, 48 hours after the SST peaks on saturday? Based on these observations, I expect significant spinup on jday 273 by 12:00 UTC. Well, since this is a prediction game, I will finish with the caveat that I have averaged about 50% accuracy on these things, so far! -Mike} ---- &color(green){9/24/08 UTC 15:00- &bold(){tcs047} is of interest today. Well, I decided to end &bold(){tcs047}. I keep changing my opinion about this, but I think I this makes sense. (Lemme know if it doesn't). When a tcs and and (invest or named storm) are not well defined to the degree that the tcs could be considered a seperate entity (as sometimes happens at higher latitudes in the eastern half of the domain) we should keep the tcs track going, even after the invest has been named. However, if the tcs and invest correspond exactly, as must be true for a typhoon for example, then we should END the tcs since it is redundant. &bold(){tcs047} is now &bold(){19wjangmi}, (was &bold(){19wnineteen}). -Mike} ---- &color(blue){9/23/08 UTC 21:12- I mended the tracks on &bold(){tcs047/96winvest04} and &bold(){tcs048/97winvest04}, particularly the gfs 54 track. &bold(){96winvest04} is showing promising signs of spinup. Circulation is increasing, we can see low-level entrainment in the fnl and gfs data, and relative wind is currently very low. Additionally, the storm has just passed the 140 Longitude mark and is making a slightly northerly turn, which has panned out well for storms in the past (most recently &bold(){15wsinlaku}). However, reality has been steadfast in defying my predictions, so nothing may come of it. Nonetheless, keep an eye out on this storm. Thats all until Friday! -Anna} ---- &color(green){9/23/08 UTC 15:15- &bold(){tcs047} is of interest today. Patka froze overnight. This is bad, since the new network wall interface was supposed to prevent that. Hmm. There are two new invests today: &bold(){96winvest04}, related to &bold(){tcs047}, SW of Guam; and &bold(){97winvest04}, related to &bold(){tcs048}, SE of Guam. &bold(){tcs046} is repaired. I just truncated the track before the moment when the navy tracking system decided the storm hyperspaced hundreds of miles to the SW! &bold(){96winvest} might be the next big storm. It has lots of moisture, low relative wind, and the SST is peaking right now. -Mike} ---- &color(green){9/22/08 UTC 17:42- &bold(){tcs047} is of interest today. I ended and deleted a lot today. tcs046 is kind of messed up since it is defined as two seperate storms by the navy. I didn't have time today to find a good solution for it. -Mike} ---- &color(blue){9/19/08 UTC 22:15- Today I touched up tracks on &bold(){tcs045} and &bold(){tcs046} and overhauled the tracks on &bold(){tcs044} and &bold(){94winvest04}. I noticed some peculiar behavior in the plots for &bold(){tcs046}. There was a very strong amount of high-altitude detrainment followed closely by very strong circulation around 250 hPa. This behavior occured near the beginning of the storm's life, around jday 259-260. Mike and I looked into the possible cause or effects but were unable to guess what might be happening here. Any ideas? Finally, I purged &bold(){93winvest05} since it has evolved into &bold(){18whagupit} and updated the September storm summary. Have a good weekend, everyone! -Anna} ---- &color(green){9/19/08 UTC 17:39- &bold(){18whagupit} is of interest today. This storm is slowly spinning up, but the sharp increase in forecasted relative winds might put the brakes on it (though, these could be due to a tracking error). I didn't have time to do any more than this. One interesting thing is that, in the sinlaku plots, the SST becomes profoundly coupled to the circulation after spinup. This is also true for fengshen, kalmaegi and fung-wong. -Mike} ---- &color(green){9/18/08 UTC 15:00- &bold(){93winvest05} is of interest today. &bold(){93winvest05} looks like it has some spinup potential. The SST seems to be peaking (and very high, at 30.5C). Also, gfs forecasts show diminished relative winds and more sporadic entrainment over the next day. Curiously, however, they do not indicate a great increase in circulation, although the zero ciculation tendency is consistent with this status. If the SST really is peaking today, I expect significant spinup to occur in 1-2 days, provided the moisture increases as predicted by gfs 54. -Mike} ---- &color(green){9/17/08 UTC 15:00- &bold(){93winvest05} is of interest today. &bold(){15wsinlaku} is spinning up &bold(){again}! The gfs forecasts a huge rise in circulation tendency over the next couple of days, but that comes with a huge drop in SST. I get errors on tcstcanal with &bold(){tcs045} fnl and &bold(){tcs047} gfs 30 and 54. Everything else is roughly tracked. -Mike} ---- &color(blue){9/16/08 UTC 20:30- I did not perform most of my tidying-up tasks today since I have come down with a head-cold and do not wish to stay long in the lab. However, since &bold(){93winvest05} is of interest today, I tightened up the GFS tracks in hopes that I would see more signs of spin-up. However, GFS is still predicting that the storm will begin to lose power within the next 24 hours. Other than that, the website looks pretty clean. I will hopefully be feeling better by Friday and will work longer then. Feel free to email me if you need anything done in the meantime! -Anna} ---- &color(green){9/16/08 UTC 14:52- &bold(){93winvest05} is of interest today. The invest is just north of Guam and is showing signs of increased activity, although the low level relative winds are increasing. Even though the fnl plots show lots of entrainment today, the gfs forecasts show that it will die off quickly. The gfs 54 suggests the entrainment will sputter a bit for the next couple days, and forecasts the SST to rise quite a bit (over 30 C!). The rainrate is expected to rise, too. -Mike} ---- &color(green){9/15/08 UTC 21:39- &bold(){15wsinlaku} is of interest today. Just a quick tracking of all storms, today. We had a brown-out and trouble with Heron. That stuff is fixed for the time-being. There is a lot that needs to be touched up for sure! My tracks are pretty messy. Also, I didn't have time to make a monthly time plot for today. -Mike} ---- &color(blue){9/12/08 UTC 20:30- Today I completed a lot of housekeeping chores around the website. I purged old storms &bold(){98winvest04, 99winvest04, 95winvest03, 15wfifteen, 037tcs035,} and &bold(){97winvest03}. Most of these were either doubles of existing storms or damaged data. I fully equated &bold(){99winvest04} to the new storm &bold(){91winvest03} and ended its GFS-54 track as it exitted our domain. I also touched up the GFS-tracks on &bold(){16wsixteen}. Finally, the storm-summary plot for September is finally up! On a side-note, &bold(){91winvest03} is showing a substantial amount of lower-level entrainment, especially in the GFS-tracks. However, due to its position and direction of motion, I predict that it will not spin up. Unfortunately, it is leaving our domain so it will be hard to follow up on my theory! -Anna} ---- &color(green){9/12/08 UTC 15:30- &bold(){15wsinlaku} is of interest today. I ended &bold(){99winvest04}, since it seems to have been renamed to &bold(){91winvest03}. The &bold(){043tcs043} track is kind of iffy at this point. Anna, you might have a better idea of what is happening there. -Mike} ---- &color(green){9/11/08 UTC 14:30- &bold(){15wsinlaku} is of interest today. Got similar processing error on &bold(){tcs038} and &bold(){tcs042} fnl today. &bold(){tcs039} still looks to be spinning up. That notion is strengthened by the invest named &bold(){99winvest04} in the same location. The tracks are now collocated, but &bold(){tcs039} shows more of the development history. In particular, a peak in SST precedes spinup by 24-48 hrs, a pattern seen also in &bold(){037tcs037} (&bold(){16wsixteen}), &bold(){017tcs017} (&bold(){96winvest02}) and &bold(){015tcs015} (&bold(){13wnuri}). GFS 54 for &bold(){tcs042} continues to indicate rising SST on its approach to Guam, though, the water content is dropping. The fnl plot shows that it *is* heading NW, but it is hard to tell anything more than that, since the plots are messed up. -Mike} ---- &color(green){9/10/08 UTC 15:30- &bold(){15wsinlaku} is of interest today. Touched up everything today. Hmm. &bold(){tcs039} looks like it might be doing something. There is some convergence in that area. If it is wet air coming in, then maybe that is what is driving the sharp increase in sat. frac. over the last 24h in the gfs 30 plots. Also, there is a hint of entrainment at low levels, and the SST recently at 30 C. Also, gfs 54 for &bold(){tcs042} indicates very high SST on the approach to Guam. Lets see if it really *does* head NW, as the model suggests. That one is a little tricky to track. There is a problem with &bold(){tcs041}, gfs 30. It woudn't process past today at 06Z. Then, fnl did the same thing for the same storm, but the 30 gfs went back to normal. Weird bug... Now, it is &bold(){tcs042} gfs 30 that is giving me the same trouble. I tried to touch up &bold(){98winvest04} a bit. Then I copied it into 16wsixteen, since they are the same. Then, I truncated 98winvest04 and ended it. Anna, you may want to retrack 16wsixteen if you still think that track has issues. The domain images are pretty jumpy for the gfs model, though. Also, &bold(){99winvest04}, gfs 54, probably needs some work -Mike} ---- &color(blue){9/09/08 UTC 18:20- I completed some quick and simple track updates today: &bold(){tcs041} got a GFS-track overhaul and &bold(){tcs033, 037, 038, 039, 042} were all touched up (particularly the GFS tracks). I also made some ammendments to the GFS track on &bold(){15wsinlaku} although it could use some more touching up. Finally, &bold(){98winvest04} needs a GFS-track overhaul as well, but I didn't get a chance to complete it. If anyone has time, feel free to get it working, but if not, I will fix it on Friday. See you then! -Anna} ---- &color(green){9/09/08 UTC 15:30- &bold(){tcs037} and &bold(){15wsinlaku} are of interest today. &bold(){039tcs038} may be showing increased activity. It has had bouts of increased rainfall and vorticity for the past few days, but now there is some entrainment. The gfs forecasts, however, do not portend anything, here. &bold(){15wsinlaku} looks to be a mighty storm, and gfs indicates it will continue to spin up over the next 48 hours. In hindsight, we probably could have seen this coming. There was upper-level detrainment over the past three days, during which the SST rose to 30 degrees! -Mike} ---- &color(green){9/08/08 UTC 18:13- &bold(){tcs037} is of interest today. There is a new named storm, &bold(){15wfifteen}, and a number of invests. I have ended &bold(){97winvest03} and &bold(){96winvest03}. One is due the old navy naming issue, which has no practical solution at this point, except to disallow the storm from continuing. Since &bold(){tcs040} covers the same disturbance as &bold(){97winvest03}, I have ended the invest (also, it is moving out of our domain). As for &bold(){96winvest03}, it is simply out of our domain (to the west). &bold(){tcs037} looks to be spinning down as predicted by gfs models. I wonder if all the upper-level wind shear kept it from spinning up, even though the SST, and low level shear were optimum for storm development. Still, the overall slope on circulation is increasing, and it may spin up a little more. -Mike} ---- &color(green){9/07/08 UTC 18:27- &bold(){tcs037} is of interest today. This is adjacent to &bold(){94winvest03}. The gfs plots for that storm show spinup over the next day and then dissipation of the storm center accompanied by plummeting SST and saturation fraction. The gfs 30 forecast agrees with gfs 54 about the SST, but the predicted entrainment is far greater. I think a worthy analogue to this storm is &bold(){90winvest02}, which was also east of 140E, and had a precipitous drop in SST and sat. fraction at the time of its spindown. -Mike} ---- &color(green){9/06/08 UTC 22:46- Well, &bold(){tcs030} is ended. It finally reached the end our domain. Anna, I retracked part of gfs 30 and 54 for &bold(){tcs033}. There was a place where the streamlines diverged, and I thought the the southern route more closely matched what the Navy was calling &bold(){tcs033}. &bold(){tcs039} and &bold(){tcs040} are pretty undefined. &bold(){tcs040} has moved out of the domain, so I ended it. &bold(){tcs037} is of interest today, with Dr. Raymond flying out there right now. However, the gfs forecasts do not indicate very much activity. -Mike} ---- &color(green){9/05/08 UTC 21:30- Nice work on the colors, Anna. I decided to try out green, since that is my favorite color. (Well, that's what Saska has determined from my wardrobe, anyway). Thanks for doing those gfs tracks. I decided to truncate those because I was overwhelmed by lots of little tcs's. I'm glad that someone got around to finishing them. In particular, I think I had ENDed the gfs tracks for &bold(){tcs033}, but that one actually is being tracked by the folks at tparc, so it is important. I noticed that this wiki editor has lots of helpful hints on tweaking text effects, below. I haven't gotten around to checking them out, yet, though.) -Mike} ---- &color(blue){9/05/08 UTC 19:30- Hey, guys! I figured out how to add color to our individual comments! Let me know if you like it or if its too weird looking. Mike, I took the liberty of leaving your text black, since you're kinda the admin of the site... feel free to switch it around if you like! -Anna} ---- &color(blue){9/05/08 UTC 18:30- I am finally back to the website. Some of the recently deactivated storms appear to be missing GFS tracks such as &bold(){tcs035} and &bold(){91winvest02} and GFS for &bold(){tcs033} seems abreviated. I will try to get those worked out today. Also, I intend to purge &bold(){92winvest04} from the website since it appears to be entirely out of our domain and cannot be tracked. Just for good measure, I won't delete the data-files yet so it can be returned to the website if someone wants it. If not, I will remove it permanently next time I come in! -Anna} ---- &color(green){9/05/08 UTC 15:00- &bold(){tcs 37}is of interest today. The fnl data continues to show hints of potential to spinup. For instance, the SST is rising, the circulation tendency is fluctuating, and the relative winds are dropping. However, the signs of entrainment from yesterday have disappeared and the gfs forecasts show no signs of future activity. -Mike} ---- &color(green){9/04/08 UTC 16:00- all tracking done for this morning. &bold(){tcs 37}is of interest today. The fnl data shows a little low level entrainment, but the gfs forecasts don't show much beyond that. -Mike} ---- &color(green){9/02/08 UTC 15:40- tcs 30 seems to have fizzled out. GFS 30 and 54 didn't predict anything much, either. The mission to tcs 30 was cancelled this morning. Nothing much is happening, except maybe tcs 33, which is headed for Guam. However, gfs doesn't give any indication of increased activity. Also, as Anna pointed out, storms E of 140E don't seem to spin up. -Mike} ---- &color(green){9/01/08 UTC 23:30- Website back up. ncl-convert2nc was part of the problem. I changed the calls to ncl_grib2nc. That seems to work. We'll see later on. tcs030 is of interest, now. It is right over Guam at this time. Everything has been tracked, but only roughly. A couple storms are ended, since they are out of range, or have no trackable features. -Mike} ---- &color(green){8/31/08 UTC 4:52- Website still down. It seems that bash is having trouble dealing with the order of processes. All of a sudden, it wants to process files that don't exist, yet. I've been troubleshooting this off and on all day, and only have half of it figured out. I have added sleep statements in tcsgrab_new.sh to slow down the cascade of processes long enough so that files have time to be fully created by one program before another one tries to access it. -Mike} ---- &color(green){8/30/08 UTC 19:52- Website temporarily disabled due to ncl bug. I'm not sure why this is happening, but ncl_convert2nc won't work with crontab all of a sudden (well, since this time yesterday). This is bad, since it is the first step in all of the data processing. It is the reason why none of the time series have been updated in the last 12 hours. Unfortunately, crontab is not sending me process output as it is supposed to. Later this afternoon, I will try to get over to Workman and check it out. -Mike} ---- &color(green){8/30/08 UTC 3:25- Nice work Anna! It looks like 90winvest02 is losing energy. We won't have our analyses to look at till tomorrow, since there was an issue with grabbing data today. All the data grabbed between 2pm and 8pm failed to get converted to netcdf for some reason. An unplausible, (yet somehow satisfying) explanation is that this issue was related to Kestrel's ups catching fire around 12:15pm (MDT). That data has been converted, however, and is ready to track in the morning. There is a new tcs031 to the NW of Guam. Also, tcs030 is moving W towards Guam and should be there (according to gfs 54) in 45 hours. -Mike} ---- &color(blue){8/29/08 UTC 20:00- The gfs 54 and 30 models at this point, both predict that 90winvest02 will NOT spin up as we previously assumed. The fnl data still looks promising, but if the gfs predictions are right, we can expect to see a decrease in low-level entrainment within the next 6-12 hours. I speculate that the weather patterns at the storm's current location are not conducive to storm-building. For instance, all of the "named" storms so far (with Nuri as the one notable exception) began spin-up to the West of Longitude 140. Just a thought... let me know what you think! -Anna} ---- &color(green){there seems to be bad data causing a bug in the analysis for tcs 27, fnl. However, the other analyses are up to date and mostly bug-free as of 14:52 UTC (8:52AM, MDT). 90winvest02 continues to exhibit factors predicting growth: low-level entrainment continues, in concert with a rising pocket of circulation at 850mb. SST and sat. frac. remain high and, perhaps most significant, the relative winds are dropping. -Mike } ---- &color(green){TCS UPDATE: tcs025 is now called 90winvest02. Previously, it was called 99winvest03. 99winvest03 is now purged from the site, and we are only tracking 90winvest02, which retains the full history of tcs25. That storm has potential to increase circulation and there is a mission scheduled today for 22:00 UTC to investigate. The fnl analysis shows lots of entrainment, optimal sst, and a recent dip in the relative wind, all factors which predict increased activity. Furthermore, the gfs 30 hour forecast agrees. -Mike} ---- Hello Everyone. Here is a central location for us to share ideas and information. Just click 'Edit', above and edit this page when you want to add a comment or note. NOTE: four dashes '----' makes a nice grey line below your entry. Use this backbutton to return to our homepage: ***[[TCS08_Project>http://patka.nmt.edu]] Have fun! -Mike ---- Note: Today, NOAA's sst ftp site is down. That caused trouble with timeplot rendering, since the sst could not be included in the analyses. We are downloading sst data every monday evening. The timeplots are fixed; however, we should be aware of the possibility that NOAA's site may be down next monday, which will cause blank areas to show up in right sides of all the time series plots. If that is the case, we will have to search for a backup copy of the sst data, which may be available on one of the NOMAD servers. -Mike (tuesday, 18:08, UTC) ---- The NOAA sst site is back up. Current focus is 099winvest03, which was 025tcs025. There is a mission scheduled today for 18:00 MDT, 22:00 UTC. -Mike ----
[ ADSPACE ]