Hand-drawing Weather Map & Forecasting Tips for Newbies

Greetings all,
This may be a biiiiiiig set of questions, but I'm hoping to hit a rich vein of veteran stormchaser's knowledge by posting them here all at once. Feel free to answer any part, if you don't feel like responding to all of the questions. Thanks in advance to those who share their approach!

I'm reading (and agreeing with) the reasons for not simply relying on software generated maps when making severe storm forecasting. Foremost among these reasons is that it slows one down to really think about the various forces/ingredients at work at the various levels of the atmosphere. I've taken a General Studies course on the Atmosphere, so I understand the basics of wind bars, drawing isobars, isotherms, etc.

I'd like to know more about other's approaches to this process:

a) Am I correct in thinking that this is only useful with actual data for current conditions (not model projections?) Or do you only look at actual data only on chase day and rely on the models prior to that?

B) Do you write in the individual station readings on a blank map, or start with a map that has those readings?

c) Is there a good online source for good-sized nicely printed maps (either blank or with station readings)?

d) What things do you particularly look for (and how do you draw it in) on each level of the atmosphere?

e) On chase day, before convection starts firing (assuming you are in your general target area) what are the final parameters that you check online before heading out?

Darren Addy
Kearney, NE
Over reliance on model output...

Yes, if you depend exclusively on model output, you're going to be limited in your success. Take last Sunday (4/10/05) in KS, for example. Anyone who chased exclusively by where the SPC mesoanalysis severe parameters lined up (CAPE's, Sig. TOR, supercell index, etc) in south central KS didn't see much. That day, the key to tornadogenesis was small-scale regions where winds backed just N of the dryline/cold front, causing locally increased values of (directional) shear and helicity. This was not well sampled and ingested into the models - anyone who was looking at RUC data that afternoon will recall very low values of SRH's in those areas, which along with limited instability suggested limited tornado and severe potential. The only way to accurately assess what was going on was to carefully watch surface obs...

This site is my personal favorite for surface obs...

As a side note, visible satellite gives a wealth of insight into the pre/early storm initiation environment. Careful observation of high-resolution satellite give clues as to quality of shear and CIN.

My favorite site for satellite imagry:

On to another subject, my AM storm day hand analysis procedure:

Every morning before heading out to work, I do a "hand analysis" of 12Z surface and upper air data on active weather days. I start with the upper-air charts. I print the unanalyzed maps from the SPC site on 11x17 sheets.


I start with 250 mb – near the tropopause – about 34,000 ft above sea
level. I should probably use 200 mb in the summer, but the SPC site only has 250 mb unanalyzed maps. I draw height contours every 120m in black – or sometimes I do streamlines instead – they should line up with the height contours very well at this level. Next, isotachs – lines of constant wind velocity. Usually, I do these in different shades of blue and purple at isotachs appropriate for the particular map – for example, > 50kts is light blue, >75 kts is dark blue, and >100 kts purple shading. I note long-wave trough axis using a thick black line, and ridge axis using a thick black zig-zag line. Look for jet streaks – regions of strongest wind flow in the jet stream. There are two principal reasons for analyzing the 250 mb (or 200 mb) levels: 1) locating areas of divergence and 2) this level, in addition to the other levels down the surface are used for assessing the shear environment. For "straight jet streaks", divergence is
often maximized near the location of the left-exit and right entrance regions. Finally, take a careful look at temperatures, as isolated areas of relatively warm air at this level can indicate divergence. With practice, it should take about 3-5 minutes to do this chart
and the other upper charts that follow.

500 mb level (about 18,000 ft). I do a similar analysis – this time
drawing heights at 60m contours. Do isotherms at 5C contours.
Important features at 500mb: areas of positive vorticity advection
(PVA) and thermal fields. Note troughs (long wave and short wave
– short waves are often better resolved at 700mb) with the same black
lines along the axis that are used at 250mb. I put a solid black
"X" at areas where PVA is maximized. These can be correlated with areas
of large-scale lift and surface convergence = best place for storms at the surface.

700 mb level (about 10,000 ft). At this level, a careful temperature, moisture (or lack of), and wind-field analysis is important. Areas of PVA are also assessed. Draw height contours at every 30 m. Isotherms and isodrotherms at 5C contours. I may also do dewpoint depression isotherms as an alternative to isodrotherms. Draw fronts aloft using standard notation – don't shade barbs and pips because this is an upper-air chart. Locate lows and trough axes. Important: note areas of dry intrusions – where lower dewpoints / increasing T/Td depressions are moving into.

850 mb level (about 5000 ft). This level is where the low-level jet
(LLJ) lives. This relatively narrow band of stronger winds out of
the south is often found in the warm sector ahead of advancing
cyclones. The LLJ winds often increase significantly after sunset
after the boundary layer decouples and the LLJ winds can
"glide" over
the boundary layer inversion without friction. At the 850mb level,
draw isotherms and isodrotherms at 5C contours. Note thermal and
moisture axis and ridges. Draw heights at 30m contours. Note
fronts, lows, and troughs; and mark using the same notation as for
the other levels.

Surface: By now, the region of the chase target should be clear – at least down to the state level. Print out a regional 12Z (or later if
available) and carefully notes isobars, isotherms, isodrotherms,
fronts, lows, mesoscale features such as outflow boundaries. Note
such things as existing clouds and precipitation…

- bill
Unanalyzed Charts are nice to use, they just include the plotted data.

Upper Air Plotted Data includes:
Dew Point Depression
Geopotential Height
12 hr Height Change
Wind Direction
Wind Speed

First off, I look to see if any suspicious data sticks out
and circle that with a regular pencil. Also by examing
current data for clues whats going on at the present
and looking at the past 12 and 24 hr charts
for continuity is a good pre-analysis routine.

From that you can contour the following isopleths,
I usually sketched it lightly with a pencil, before using color pencils.

Isallohypse: Contour in meters at 500 mb and 700 mb
Isodrosotherms: Contour interval of 2 deg C

Isohypse: Contour interval of 120 m at 250 or 300 mb
Contour interval of 60 m at 500 mb
Contour interval of 30 m at 700 mb, 850 mb and 925 mb.

Isotachs: Contour interval of 20 kts at 250 or300 mb
I will vary the interval at the other levels to my liking.

Isotherms: Contour interval of 2 deg C or 5 deg C

Color Pencils I use when making a composite chart:
250 or 300 mb: Purple
500 mb: Blue
700 mb: Brown
850 mb: Red and Green

What to look for on the charts: I know I am missing a few things:

250 or 300 mb chart:
-Height Troughs and Ridges
-Highs and Low Centers
-Jet Steaks
-Significant Speed Shear
-Significant Streamlines

500 mb chart
-Cold Pools
-Diffluence Axis
-Height Troughs and Ridges
-Highs and Lows Centers
-Jet Streaks
-Postive and Negative Vorticity Advection (Get good idea where they are)
-Significant 12 Hour Height Falls
-Significant Moisture
-Significant Streamlines
-Significant Temperature Falls
-Speed Shear Zones
-Thermal Troughs (Cold)
-Vorticity maximums (you can get a good idea where they are)
-Warm and Cold Air Advection

700 mb
-Diffluence Axis
-Dry Line
-Dry Intrusion
-Highs and Low Centers
-Jet Streaks
-Short Waves Troughs & Ridges
-Significant Height Falls
-Significant Moisture
-Significant Streamlines
-Significant Temerature No Change Line
-Significant Temperature Falls
-Thermal Ridges (Warm)
-Thermal Trough (Cold)
-Upper Level Fronts
-Warm and Cold Air Advection
-Warm and Cold Pockets

850 mb (Also 925mb)
-Confluence and/or Convergence
-Height Troughs amd Ridges
-Highs and Low Centers
-Jet Streaks
-Maximum Moisture Axis
-Significant Moisture
-Significant Streamlines
-Thermal Ridges
-Upper Level Fronts
-Warm and Cold Air Advection (WAA)

One more note, you can use these charts with
current weather data and compare it to the initiaization
run of the different models to get a general idea if
the models pick up the various features.

Drawing skew-t s by hand can be time consuming,
decoding TTAA data and then plotting it on your
laminated Skew-T and then calculate a lot of parameters
from it, plus you can make your own hodographs.

I usually look at the soundings and forecast soundings
from various web sites.

More on hand analysis:
also Amos had a thread on hand analysis, cannot find the link.

Well, it's probably just because I'm lazy - but I don't really find time to do as in depth a hand analysis as that described above. While the detail described above is thorough enough to then make real assessments of the quality of the model initialization fields, I prefer to focus on the variables that I consider most important to anticipate the synoptic environment that will be present later in the day. I also use the unanalyzed maps from SPC, I really like the pdf format for when I want to do digital overlay using a software package. Anyhow, here is my abbreviated list of analysis:

250 mb - contour isotachs only, and roughly line in the jet axes. I then look for regions of notable diffluence/divergence and encircle and lightly shade.

500 mb - contour of heights, then identify jet streaks/cold pools (positive and negative vorticity maxima) - I almost always use a water vapor loop in order to precisely find locations of features - and then draw an arrow pointing in the direction of motion of these features.

700 mb - contours of temperature, and shade in regions of high RH, cold and warm advection. Draw in the axes of warmest and coldest temperatures at 700 mb. Overlay fronts if identifiable.

850 mb - again contours of temperature - but also contours of dewpoint. As this is a mess west of the Rockies - I only do this for east of the Rockies. Again, mark the warm/cold axes as well as moist axis, identify any fronts, warm and cold advection regions shaded. Mark in the axis if the 850 mb jet in the warm sector.

If limited moisture is present at 850 in the warm sector - then I'll check some soundings in the LLJ to see if better moisture is lower down. If so, then I'll also repeat the analysis at 850 for 925 mb.

Surface - East of the Rockies contours of pressure, temperature, dewpoint, mark in fronts & boundaries, high/low pressure centers, cloudy regions and morning precipitation, overlay moist axis. Bring up a animation of surface obs, identify the mean motion of the surface low and draw in an arrow.

Now, I try to synthesize the above information to get a feel for where things are at and where they might be in the near future to hopefully narrow down a target region about the size of the state of Connecticut.

Okay. So how does one go about actually doodeling all this stuff? I've yet to find any useful information on the internet. Also, is it possible to find data you can print on standard 8.5 x 11 paper and have it not look like a jumbled mess?
Okay. So how does one go about actually doodeling all this stuff? I've yet to find any useful information on the internet. Also, is it possible to find data you can print on standard 8.5 x 11 paper and have it not look like a jumbled mess?

Some links to maps were given above by BIll. If you are asking how to do contour analysis - yes - as you would expect there are web resources on how to do that. At the top of google I found this: