(1) Study surface analysis charts: CHART 1, CHART 2

Take note of:

a) Thermal advection (remember thermal advection is a function of the thermal gradient, the wind speed through the thermal gradient, and the angle the wind is through the temperature gradient)
b) Keep in mind the soil temperatures, cloud cover / lack of cloud cover and snow cover. These three parameters will modify air temperature as an air mass advects.
c) Wind speed/ direction
d) Pressure pattern

(2) Study 850 mb analysis chart. Take note of same items as above.

(3) Study surface and 850 mb model output (NAM / GFS) for these same items above

(4) Examine LATEST SOUNDING. Ask yourself how the sounding will be modified throughout the day. An approximation of the afternoon high can be made by taking the DALR from the top of the PBL to the surface. This works best on clear, moderate wind days in a barotropic atmosphere.


*Thermal advection (strength of WAA and CAA)

*Local heating/ cooling (ISR, OLR)

*Daylength (less daylight = less ISR = lower temperatures)

*Atmospheric albedo (i.e. clouds, snow cover, suspended particles)

*Dewpoint (sets a quasi-limit for overnight cooling in a barotropic environment)

*Latent heat release or absorption (i.e. wet bulb cooling, latent heat release from dew)

*Vertical temperature movement (adiabatic ascent, descent; downdraft from storm)

*Wind speed (determines how much heat will build at surface by day and how much cool air will build at night; strong winds keep the air well mixed)

*Mesoscale effects (i.e. topography, urban heat island, water bodies)

High and low temperatures are a primary function of thermal advection, wind speed, cloud cover, dewpoint and the number of daylight hours.

*Highs can be less than expected due to CAA, high wind speed, increased cloud cover, higher dewpoint and shorter daylength.

*Highs can be greater than expected due to WAA, low wind speed, decreased cloud cover, lower dewpoint and longer daylength.

*Lows can be less than expected due to CAA, low wind speed, decreased cloud cover, lower dewpoint and longer nights.

*Lows can be greater than expected due to WAA, high wind speed, increased cloud cover, higher dewpoint and shorter nights.

*Mesoscale effects such as urban heat islands, differential vegetation, topography, nearby lakes / rivers / oceans and altitude must also be taken into consideration!

*Low wind speed on sunny days will result in warmer temperatures than if the winds were stronger. With light winds, heat can build right at the surface without being significantly mixed with cooler air aloft. This can form what is known as the superadiabatic lapse rate. This is the opposite case on a clear night. Light wind at night does not allow radiationally cooled air at surface to mix with warmer air aloft.


Fronts may cause highs and lows to occur at untraditional times during the day. In association with a strong cold front, the high will occur before frontal passage and the low will occur at midnight (assuming CAA and temperatures continue to decrease throughout the day). The timing of the front is critical in determining what the high will be before the front passes and how much CAA will occur when the forecast period ends (usually midnight).


Sloped terrain produces downsloping and upsloping wind especially when the wind direction is perpendicular to the slope of the terrain. A wind direction forecast is critical in determining how much adiabatic warming or cooling will occur along the slope. Upslope cooling in the cool season can produce adiabatic cooling and snow while downslope flow produces adiabatic warming and a chinook wind.


Wet-bulb cooling will occur over regions precipitation occurs. If the precipitation forecast is incorrect, odds are the temperature forecast will suffer also. Afternoon thunderstorms can cause the high temperature to be cooler than predicted. Any rain during any time of the day will cause evaporational cooling at the surface. The low level dewpoint depression determines how much surface temperatures will cool. A high dewpoint depression will result in a greater evaporational cooling. Convective thunderstorms also transports air from higher in the atmosphere to the surface. This will alter surface temperatures greatly in these downdraft regions.


*Temperature accuracy diminishes rapidly after 3 days

*Use GFSx extended forecast temperatures and your knowledge of the JET STREAM to make extended temperature forecast

*Underneath trough--- cooler than normal

*Underneath ridge--- warmer than normal

(5) Study GFS / NAM MOS data and METEOGRAMS after coming to your own forecast consensus. Tweak forecast if data is reading a process you did not consider. Learn the MOS, meteogram biases and PITFALLS for your forecast region.

(6) Read NWS zones and state forecast discussions

(7) Debate in your mind why you agree or disagree with the NWS and computer models.

(8) Make your temperature forecast (lows, highs). Keep in mind the starting and ending time of your forecast period. Note also how the forecast verifies. This will prevent the same forecasting mistake from being made in the future. Experience adds greatly to making correct temperature forecasts.