The Diurnal Temperature Smart Tool

1
The Diurnal Temperature Smart Tool
Angie Enyedi
National Weather Service
Jacksonville, FL

• What is it and why the need for it?
• How is it implemented?
• What are the results?

2
What is it?
An interpolation method of hourly temperatures
between forecast maximum and minimum temperatures
in the new graphical forecast produced by the
NWS, the Interactive Forecast Preparation System
(IFPS).
3
Initial Interpolation Techniques

• The default interpolation methods in the
  Graphical Forecast Editor (GFE) were either the
  Cubic Spline or the Linear Interpolation Methods.
  
• These methods were used during the
  implementation stage of IFPS.

4

• Both fail to represent abrupt temperature falls
  after sunset and temperature rises after sunrise.
  
• Both have unrepresentative peaks (linear) and
  plateaus (cubic spline) when the daily max or min
  temperature is achieved.

5
We need an interpolation tool that accounts for
monthly differences in radiational cooling and
insolation between the times of daily minimum and
maximum temperatures! ANSWER The Diurnal
Temperature Tool
6
(No Transcript)
7
Cubic Linear Diurnal MAE (? F) 4.7
3.4 1.1 Mean Absolute Error
8
How the Tool was Implemented at JAX1.
Stratifying the Data

• 7 years of hourly ASOS observations were
  stratified by month then by hour for site JAX
  (Jacksonville, FL).
• Archived data was obtained from NCDC.

Date/Time (Z) Actual Temp Apr 1, 1995
0200Z 55 Apr 2, 1995 0200Z 53 .etc Apr
30, 2001 0200Z 66
9
2. Averaging the Data
All temps in a particular hour in a given month
were avearged to get an average Thour.
All Temps for that hour summed, then divided by
total number of entries to get Thour. Average
Thour for 02Z in April 59.56
For Example Thour for 01 April 60.30 Thour
for 02 April 59.46 Thour for 03 April 58.59
10
3. Calculating the ratio, f
Temperatures from 00Z-23Z were converted to a
range between 5-95 based on the difference
between the Tmax and Tmin each month.
Note f for Tmax 95
f for Tmin 5 This was done to reduce the warm
and cold bias.
(Thour-Tmin) Tmax-Tmin 5
f
90

11
4. Smart Tool references the f Table
F Table for April 00Z 58 12Z 12 01Z
45 13Z 33 02Z 38 14Z 53 03Z
33 15Z 68 04Z 27 16Z 78 05Z
23 17Z 85 06Z 20 18Z 91 07Z
16 19Z 94 08Z 12 20Z 95 09Z
9 21Z 92 10Z 7 22Z 86 11Z
5 23Z 75

• The f table is referenced by the diurnal
  temperature smart tool when it calculates hourly
  temperatures (Thour) between the forecast daily
  minimum and maximum temperatures.
• Tool will warm cool temperatures based on
  monthly diurnal climatological trends.

12
(No Transcript)
13

• Note the seasonal variation of the approximate
  time of the daily max and min temperature.
• Note the seasonal variation of temperature
  decrease (increase) after sunset (sunrise).

14
Cubic Linear Diurnal MAE (? F) 6.4
3.8 1.7 Mean Absolute Error
15
Cubic Linear Diurnal MAE (? F) 4.4
3.1 1.8 Mean Absolute Error
16
Cubic Linear Diurnal MAE (? F) 5.1
3.1 0.8 Mean Absolute Error
17
Results of the Diurnal Temperature Interpolation
Method
Data suggests that the diurnal temperature
interpolation method better represents monthly
radiational cooling insolation trends when
compared to the linear and cubic spline methods.
18
Benefits of the Diurnal Temperature Interpolation
Tool

• Best represents abrupt diurnal temperature rises
  and falls when compared to the cubic spline and
  linear interpolation techniques.
• Has representative increases and decreases when
  the daily max or min temperature is achieved.
• Can be customized for each month (and various
  sites across the country) accounting for
  differences in radiational warming cooling.

19
Short Comings of the Diurnal Temperature Smart
Tool

• Only customized for site JAX in our forecast
  area. The tool is run over our entire grid space
  and may not represent radiational cooling or
  insolation trends at sites further inland nor
  those close to the coast.
• Does not represent sensible heat influences from
  convection or latent heat influence from ground
  moisture.
• Does not account for atypical warming/cooling
  such as warm air advection at night or cold air
  advection during the day.

20
Taking the Diurnal Smart Tool a Step Further
Data could be further stratified based on cloud
cover, for example Diurnal temperature smart
tool could be created for sunny/mostly sunny,
partly cloudy, and clouldy/mostly cloudy days in
GFE.
21
Credits Where to Get the Tool

• Cheryl Sharpe and Steve Nelson (NWS)
• Created the original Diurnal Temperature Smart
  Tool
• Jason Hess
• IFPS Focal Point NWS Jacksonville, FL
• The Diurnal Temperature Smart Tool is available
  for download into IFPS at the Smart Tool
  Repository Website.

22
Thank You!

• Questions and/or Comments?

Angie Enyedi
Forecaster NWS
Jacksonville, Climate F.P.
angela.enyedi_at_noaa.gov