How many chasers make their own forecasts?

[Robert Dewey]

I rely solely on TWC's TORCON. Okay, just kidding... but lump me in with the other 99% that make their own forecast. I'm a total weather-geek all around. I love looking at maps, graphs, charts, and anything to do with data. I'll admit that I rely heavily on NWP, but I definitely don't stick to just one model in particular. While I tend to favor the RUC / NAM / HRRR on the "day of," I also look at the Canadian NWP suite, GFS, and SREF 1-3 days before an event. The standard nowcasting data consists of a combination of RUC / HRRR and satellite, radar, and surface data... including SPC mesoanalysis. This helps me see what's going on in real-time, and how things will evolve over the next several hours.

To be honest, forecasting really is an art-form. It's not something they can teach you as there are no "text book" repeats in meteorology. Experience is probably the single most important factor, which when combined with statistics (i.e. knowing that 7 out of 10 times the NAM is too deep over a particular area) can really enhance your skill.

 

[rdale]

To be honest, forecasting really is an art-form. It's not something they can teach you as there are no "text book" repeats in meteorology. Experience is probably the single most important factor, which when combined with statistics

eh, gonna pull a "what?" on that one... If all you do to make your forecast is model reading, then I agree. However people can learn the detailed inner workings of the atmosphere and even make forecasts without models - which combined with experience will do much better than just adding statistics to a NAM perusal. I hear some people even get paid to teach that aspect

 

[Robert Dewey]

eh, gonna pull a "what?" on that one... If all you do to make your forecast is model reading, then I agree. However people can learn the detailed inner workings of the atmosphere and even make forecasts without models - which combined with experience will do much better than just adding statistics to a NAM perusal. I hear some people even get paid to teach that aspect

I'm not sure what you don't understand about it. You've obviously read too much into my short example of the NAM. Perhaps I should preface it with "ex:" instead of "i.e." next time

My subtle implication was actually quite the opposite of what you seemed to get out of my statement - the fact that a forecast is about much more than looking at a map and saying "yep, X marks the spot." Climatology and statistics / probability are a huge factor, and don't necessarily apply strictly to model data (although it can, as per my previous example). Everything from narrowing down a "most likely" region where storms will "probably" occur, to assessing the most climatological time for tornadoes to occur, or using your past knowledge that 250j/kg of CAPE with 80knts of bulk shear sweeping along a cold front through MI means trouble.

I hope that clears it up for anyone else.

http://www.cimms.ou.edu/~doswell/probability/Probability.html
http://www.hpc.ncep.noaa.gov/html/model2.shtml

 

[Matthew Piechota]

I know when I first started forecasting I was looking alot at SPC and the risk area and such which Im sure alot if not most chasers still glance at what the SPC is doing, but now after I have been doing my own hand anaylsis ever morning of a chase (all UA and a sfc map and also a composite map) I will pick my target based on my maps. There has been nothing more helpful to my forecasting skills than doing hand analysis over and over again for every setup. contouring the isotherms/isotachs and such really has helped me understand and forecaster better over the last few years. There are days when the target is clear as day to see but sometimes its not and doing it yourself can help you find those subtle targets.

There are also alot of other things out there to help you with a forecast like the SREF/HRRR which some have mentioned. I know the HRRR has been used lately for CI and sometimes its dead on and other times its not but there is nothing better than doing the analysis yourself or a sfc map every hour up until storms go and sorting out the mesoscale details of whats going on.

 

[Verne Carlson]

I only use the SPC outlooks if I get slammed at work and don't have time to look at the models. I will do a dive into the models before going on an overnight trip in all cases though.

You have to realize that the SPC outlook is for the entire 24 hour period, not just the peak heating tornado threat. It may be a wind, hail or derecho threat that is prompting that moderate or high while the tornado risk is low. Something to factor in.

A browse at the WFO discussions and a look at the 4km WRF sim radar runs is a must as well.

 

[Corey Sloan]

I also do most of my own chase forecasting...I developed a flow chart that I use to look at everything from upper air to surface obs to MOS guidance for consistency in my amateur forecasting....I do look at the SPC and AFD's as the last item on my flow chart to see if I am anywhere close, like Steve said, to the folks who are mets and get paid to forecast....I also try to document my thoughts in a blog to see how close I was or how ignorant I was on the setup I also go back and pull the data from severe events to see where I made my mistakes.....the web is full of training resources on forecasting and weather related items.....

If it is a chase close by (I.e. after work chase) and I have been busy at work it is SWODY1, AFD's, MD's, surface features, RUC, 4km WRF and HRRR as I head out the door...I normally have had a chance to peruse the night before model data (usually 0Z) but not usually the day of model data in this type of scenario....

 

[rdale]

I'm not sure what you don't understand about it.

The part about not being able to teach people how to forecast...

 

[Terry Tyler]

Based on this thread, I make my own forecasts; but they arent my forecasts. They belong to the NAM, GFS, RUC, etc. Those computer models do all the 'forecast work' for me. The only work I do is knowing how to interpret the indicies. Anyone who messes with them know it doesnt always work out how the run put it. The models people 'make their own forecasts' with make the forecast themselves. Those people; just like me, only follow the output of whichever model that does the 'forecast'. With all the modern technology available, it really is just common sense when you know the basics of meteorology. Even to a layman off the street; If you look at the run and it says that "X" indicie is high in one area and not another. You can assume that the higher area will be at better chance of something happeneing.

I guess the arguement here roots from the 'pride' associated with 'making ones own forecast'. It comes from the weather buffs and it isnt such a bad thing. Knowledge, understanding, and respect of the meteorological components, indicies, and ingredients that 'make' a 'forecast' help give better appreciation and understanding for the weather behind it. Being a storm chaser, that kind of knowledge is priceless and important when it comes to consistantly making intercepts on storms. It puts you in another stack then those who dont pay attention to the details. Those trying to do this professionally arent concerned about who does 'their own forecast'.

 

[Dave Rennie]

I have, in the past, relied on SPC and NWS Nexrad radar displays during my chasing, but would like to learn more in order to try my hand at forecasting. I found two books on Amazon by Tim Vasquez and am wondering if anyone has used them and/or would recommend them: Weather Map Handbook and Weather Forecasting Handbook. Another book I found there was by George Elliot titled Weather Forecasting: Rules, Techniques and Procedures. Any comments on any of these books, or other suggestions, would be greatly appreciated! Thanks.

 

[Robert Dewey]

The part about not being able to teach people how to forecast...

Your "just adding statistics to a NAM" comment threw me off. I agree you can be taught to a certain extent, but I still say experience is king...

 

[Jeff Duda]

Given the huge amount of free data available to anyone with an internet connection, very few people truly make forecasts 100% from their own head. I am no exception to this. I look at SPC convective outlooks as well as the traditional American models (GFS and NAM), but also look at other models such as SREF, FIM, RR/HRRR, and the wide variety of high resolution WRFs. Basically I take in as much forecast/model input from as many sources as I can and generally look for a consensus that seems reasonable, then add perturbations from the consensus from what my experience/knowledge tells me. I think discovering trends in the models is a great way to increase chase success, but it's something I don't currently do with any regularity. I should get in to that.

I'm not someone who has a lot of money and/or time to just go chase anywhere I want. I generally am restricted to leaving and coming back in one day, so I do have a bit of a leash. That does impact my targeting. I generally pick a target the night before (unless it's a surprise chase or one that I wasn't planning on making until I saw updated information), but once the chase day is on and I'm on the road, I'm constantly monitoring data instead of driving to my previous target regardless. It doesn't matter how good of a target I think I picked before, if surface obs, satellite, and hourly objective analyses (i.e., SPC RUC mesoanalyses) say a different area looks good, I will not hesitate to adjust my target. I haven't chased all that much (maybe 50 chases since I started), but I haven't missed a big event yet by leaving my initial target for another one after having looked at updated observations, and I've hit some big events by adjusting from my initial target.

 

[Bob Hartig]

Rob & Rob: I think you're both right.

 

[Steve Miller]

I also do most of my own chase forecasting...I developed a flow chart that I use to look at everything from upper air to surface obs to MOS guidance for consistency in my amateur forecasting....

A written flow chart is a great idea! I wonder if each of us developed a flow chart, how similar or dissimilar would they would be? I think I'll play around with one this winter and see if it helps me stay on-track with the order in which I view maps.
Thanks for posting.

 

[ahaberlie]

I know in Wisconsin, at least, that the positioning of overnight MCSs can greatly effect the positioning of the severe weather later (duh?) Many a times the SPC would forecast an area of severe potential, only to have it washed out by elevated remnant convection and clouds. The few times that we did have tornadoes in southern WI, they were positioned along the boundary of the morning MCS... even a week or so ago, the storms were within 100 miles of the earlier thundershower activity.

This might be under further review, but you may have dryline N to S boundaries a common occurance in the southern plains, whereas in the great lakes region, W to E outflow/MCS boundaries may be more predominant. Thus, forecast soundings and storm motions could effect where you target, depending on what part of the country you are in/what the surface conditions are.

The SPC forecast at about 3pm is usually pretty good, but sometimes you need more time than that to get to your target.

One example

Day 3 outlook for July 22
http://spc.noaa.gov/products/outlook/archive/2010/day3otlk_20100720_0730.html

Day 2
http://spc.noaa.gov/products/outlook/archive/2010/day2otlk_20100721_0600_prt.gif

Day 1 tornado probabilities
http://spc.noaa.gov/products/outlook/archive/2010/day1probotlk_20100722_1200_torn_prt.gif
http://spc.noaa.gov/products/outlook/archive/2010/day1probotlk_20100722_1300_torn_prt.gif
http://spc.noaa.gov/products/outlook/archive/2010/day1otlk_20100722_2000.html

July 22nd Reports
http://spc.noaa.gov/climo/reports/100722_rpts.gif

Evolution

If you went to Davenport, IA, Peoria,IL , or Waterloo,IA after seeing the initial day one tornado % area.. you wouldn't be too happy. Overall the SPC pretty much nailed this as best as they could, but we are still talking about an incredibly large area. Nowcasting using surface data is extremely useful.

 

[Dann Cianca]

Being very inexperienced, I would consider myself to be an SPC chaser. While the SPC is a great resource, I know it's best to learn how to forecast on my own. I often see posts from others discussing their own forecasts, but I'm curious to know how many other weather enthusiasts actually make their own (accurate) forecasts.

-George

I'm not ashamed to say that I look at the SPC Convective Outlooks EVERY DAY, in-season or not. I mean, when it comes to severe weather, it would be kind of stupid to ignore the outlooks since some of the best convective forecasters in the nation actually work there and issue them. Are they perfect? ... no forecast is, but I believe it to be a tremendous resource. (For the record, I think it is funny when chasers act "better than" SPC forecasts and denounce them, only to publish the SWODY graphics on their blogs and Facebooks)

Every day, regardless of season, I also take a look at the synoptic picture of the United States, generally monitoring the water vapor loops and upper air maps. Where are the large scale features? This may seem obvious, but when you have a good idea as to what is going on up top, it makes figuring out the mesoscale and surface features much easier. I like "isochronal" analysis as well, which is basically looking at a feature over time. Take the water vapor loop for example, if you see a shortwave in the flow and can "time it" from the west coast through the flow, you can get a general idea as to where it will be come late afternoon in your target area. This can be done without models and is a useful tool in wrapping your brain around a forecast.

At the surface, it is important to have a look at moisture and boundaries. I saw my first tornado on a day that only called for isolated storms because we (myself and my chase partner) had located a weak boundary that had formed from outflow the day before. The following day, our storm hit the boundary and wham, tornado-time. It wasn't even all that apparent on the surface obs since it was not in a dense network, but we could see the cu along it. If we had not seen it the day before, we may have missed it the day of. Take a look at "yesterday's" radar and visible sat loops before you head out.

MODELS:
For long range, I mainly monitor the GFS and ECMWF but only for synoptic features. Where are the troughs/ridges and what amplitude are they? Look for consistency between the models and runs.

The week of, I start watching the NAM and GFS closely to see how it is handling the synoptic systems and watch the moisture/thetaE fields.

The day before, I'm still monitoring the NAM and GFS. I also like to check out the local area forecast discussions issued by the NWS for the area(s) I want to target. Let's just say if it's in southwest Kansas, I'm praying to see "UMSCHEID" at the bottom of the AFD.

PRECIP MODELS:
http://www.emc.ncep.noaa.gov/mmb/mpyle/cent4km/conus/00/
http://www.emc.ncep.noaa.gov/mmb/mpyle/cent4km/conus/12/
http://rapidrefresh.noaa.gov/hrrrconus/
http://www.nssl.noaa.gov/wrf/

They can be very useful, but take them with a grain of salt. I look at them specifically for regime and storm type. The models really only produce a possible scenario based on the input conditions and shouldn't be followed to the letter. All three of them have their strengths and weaknesses and all of them have had days where their solutions have panned out almost exactly as forecast.

A typical use for me would be, for example: The HRRR is showing an isolated supercell on the dryline. That's a good reason to be out somewhere no the dryline.

DAY OF AND DURING THE CHASE:
I always look for morning cloudcover. Sometimes the short term models (RUC) are a little behind on this. I like sunshine.... look for boundaries on radar. They'll often show up reasonably well in the clutter on radars that are slicing/ducting through inverted, early morning air. I am also a slave to SPC's mesoanalysis page, but watch out for bad obs because that can sometimes skew the data. I find the most useful parameter to be the Mixed-Layer CAPE, especially on days when capping is present.

I'm sure there is more, but that is the way my brain typically works when forecasting/nowcasting for a chase.

Don't be ashamed to look at SPC! There's some good info there. I have my meteorology degree and I still look at it every day!