Back to Tropical Cyclone
Observation Page | Back
to Main FAQ Page
Subject: H1) What is the Dvorak technique and
how is it used?
Contributed by Chris Landsea
The Dvorak technique is a methodology to get estimates of
tropical cyclone intensity from satellite pictures. Vern Dvorak
developed the scheme using a pattern recognition decision tree
in the early 1970s (Dvorak 1975, 1984)
.
Utilizing the current satellite picture of a tropical
cyclone, one matches the image versus a number of possible
pattern types: Curved band Pattern, Shear Pattern, Eye Pattern,
Central Dense Overcast (CDO)
Pattern, Embedded Center Pattern or Central Cold Cover Pattern.
If infrared satellite imagery is available for Eye Patterns
(generally the pattern seen for hurricanes, severe tropical
cyclones and typhoons), then the scheme utilizes the difference
between the temperature of the warm eye and the surrounding cold
cloud tops. The larger the difference, the more intense
the tropical cyclone is estimated to be. From this one gets a
"T-number" and a "Current Intensity (CI) Number". CI numbers
have been calibrated against aircraft measurements of tropical
cyclones in the Northwest Pacific and Atlantic basins. On
average, the CI numbers correspond to the following intensities:
Current Intensity Numbers
CI Number |
Maximum Sustained
One Minute Winds (kts) |
Central Pressure (mb) |
Atlantic | NW Pacific |
0.0 | <25 | ---- | ---- |
0.5 | 25 | ---- | ---- |
1.0 | 25 | ---- | ---- |
1.5 | 25 | ---- | ---- |
2.0 | 30 | 1009 | 1000 |
2.5 | 35 | 1005 | 997 |
3.0 | 45 | 1000 | 991 |
3.5 | 55 | 994 | 984 |
4.0 | 65 | 987 | 976 |
4.5 | 77 | 979 | 966 |
5.0 | 90 | 970 | 954 |
5.5 | 102 | 960 | 941 |
6.0 | 115 | 948 | 927 |
6.5 | 127 | 935 | 914 |
7.0 | 140 | 921 | 898 |
7.5 | 155 | 906 | 879 |
8.0 | 170 | 890 | 858 |
Note that this estimation of both maximum winds and central pressure
assumes that the winds and pressures are always consistent. However,
since the winds are really determined by the pressure gradient, small
tropical cyclones (like the Atlantic's Andrew in 1992, for example)
can have stronger winds for a given central pressure than a larger
tropical cyclone with the same central pressure. Thus caution is urged
in not blindly forcing tropical cyclones to "fit" the above pressure-
wind relationships. (The reason that lower pressures are given to
the Northwest Pacific tropical cyclones in comparison to the higher
pressures of the Atlantic basin tropical cyclones is because of the
difference in the background climatology. The Northwest Pacific basin
has a lower background sea level pressure field. Thus to sustain a
given pressure gradient and thus the winds, the central pressure must
accordingly be smaller in this basin.)
The errors for using the above Dvorak technique in comparison to
aircraft measurements taken in the Northwest Pacific average 10 mb with
a standard deviation of 9 mb (Martin and
Gray 1993). Atlantic tropical cyclone estimates likely
have similar errors. Thus an Atlantic hurricane that is given a
CI number of 4.5 (winds of 77 kt and pressure of 979 mb) could
in reality be anywhere from winds of 60 to 90 kt and pressures of
989 to 969 mb. These would be typical ranges to be expected; errors
could be worse. However, in the absence of other observations, the
Dvorak technique does at least provide a consistent estimate of
what the true intensity is.
While the Dvorak technique was calibrated for the Atlantic and
Northwest Pacific basin because of the aircraft reconnaissance data
ground truth, the technique has also been quite useful in other
basins that have limited observational platforms. However, at some
point it would be preferable to re-derive the Dvorak technique to
calibrate tropical cyclones with available data in the other basins.
Lastly, while the Dvorak technique is primarily designed to provide
estimates of the current intensity of the storm, a 24 h forecast of the
intensity can be obtained also by extrapolating the trend of the
CI number. Whether this methodology provides skillful forecasts is
unknown.
Back to Tropical Cyclone
Observation Page | Back
to Main FAQ Page
|