Weather Forecasting ... On-Line

Temperature Forecasting


Our purpose here is to review those factors that impact surface temperature. Upon completion of this web page you will be able to:

  1. Assess the relative impact of a several physical processes on temperature changes.
  2. Effectively evaluate and modify temperature guidance.

Diurnal Temperature Patterns

Any discussion of temperature should start with the diurnal temperature changes that we all experience. From our day-to-day living we realize that it is typically warmer during the daylight hours and cooler at night. This observation indicates that the sun is the primary controlling factor in daily temperature change. Nevertheless, other factors impact surface temperature. These factors are discussed below.

In our introductory meteorology courses we learned that the diurnal cycle has a minimum around sunrise and a maximum between 2 and 5 o'clock in the afternoon, depending upon the time of year and location. The temperature range (maximum temperature minus minimum temperature) varies from day-to-day and month-to-month. The table below shows the monthly average high and low temperatures for Kansas City, Missouri, and the average temperature range based on these normals.

Month Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Average Max 34.7 40.6 52.8 65.1 74.3 83.3 88.7 86.4 78.1 67.5 52.6 38.8
Average Min 16.7 21.8 32.6 43.8 53.9 63.1 68.2 65.7 56.9 45.7 33.6 21.9
Daily Range 18.0 18.8 20.2 21.3 20.4 20.2 20.5 20.7 21.2 21.8 19.0 16.9

For Kansas City, the smallest temperature range is during the winter when the sun angle is lowest and the hours of daylight are at a minimum. The largest range occurs during the spring and fall when sun angles are higher. It is interesting that that summer temperature ranges are slightly smaller than those of spring and fall. It appears that the more humid conditions prevalent during the summer tend to reduce the temperature range compared to spring and fall.

Review of the Surface Radiation Balance

Heat Transfer across the Air-Earth Interface
Heating   Cooling
LW radiation from earth
BL Air conduction

solar radiation
LW radiation from air
surface terrestrial radiation

Let's start with a review of the energy exchange at the earth's surface. Solar radiation enters the atmosphere where very little is absorbed by the atmosphere itself. The amount of absorption at the earth's surface depends upon surface albedo.

As the table at the right shows, fresh snow has the highest albedo and reflects the most sunlight back into space. On the other hand, grass or soil absorb considerable solar energy.

Sunlight is the primary mechanism for heating the earth's surface, which in turn, heats the adjacent boundary layer air via conduction. As a result, some knowledge of the state of the earth's surface is important to temperature forecasting.

The earth continually emits long wave (LW) radiation into the atmosphere and space. When solar radiation is not present to counteract this energy loss, it is the primary mechanism for cooling the earth's surface. This cooling then affects the boundary layer air via conduction.

The amount of long wave radiation absorbed by the atmosphere is related to the carbon dioxide and water vapor content of the atmosphere. Some of the absorbed energy is re-emitted back to the earth's surface and helps counteract the heat lose at the earth's surface.

As we discuss the primary factors affecting temperature, keep these energy relationships in mind.
Surface Albedo Values
New Snow 80-85
Old Snow 50-60
Sand 20-30
Grass 20-25
Dry Soil 15-25
Wet Soil 10
Forest 5-10

the transfer of heat through matter by molecular activity;
conduction of energy is a consequence of temperature gradient
long wave (LW) radiation
a term loosely used to distinquish radiation at wavelengths
longer than 4 micrometers, usually of terrrestrial origin
solar radiation
the electromagnetic energy emitted by the sun; most of this
energy is in the wavelength interval from 0.15 to 4 micrometers
surface albedo
the ratio of reflected flux density to incident flux density;
flux density as used here is the radiant solar energy per unit time
crossing a plane surface of unit area

Primary Factors Affecting Temperature

Described below are six factors that impact surface temperature. These discussions take a qualitative perspective rather than a quantitative approach. In some situations more than one factor may be affecting a particular time and place. Multiple factors may be reinforcing or may be opposing each other. In any case, a qualitative understanding of the physical processes that are occurring is important to the forecaster.

Dry Air versus Moist Air

We mentioned above the carbon dioxide and water content of the atmosphere absorbs and re-emits long wave radiation. If we assume that the carbon dioxide content is constant, the daily variation of this factor depends upon the water content. We can look at a couple of impacts here.

Consider a warm humid summer night versus dry desert air. Over the desert at night temperatures fall off quickly because there is little water vapor in the air to absorb the long wave radiation and re-emit this energy back to the earth's surface. As a result, the ground cools off and cools the air.

In contrast, on a warm summer night the humidity in the air absorbs the long wave terrestrial radiation and re-emits some energy back to the earth, reducing the cooling compared to the desert scenario. Typically, the more humidity present, the less the cooling. Can you recall any warm humid summer nights where it was difficult to get to sleep because of the heat and humidity?

Related to this humidity factor is the amount of cloud cover. Clouds are liquid water droplets and also absorb and re-emit the long wave radiation and, in addition, reflect solar energy. During the day, an overcast cloud layer will reflect from 70 to 80 percent of incident sunlight for thick clouds to 25 to 50 percent for thin clouds. Cloud layers reduce the amount of solar energy reaching the earth's surface. This, in turn, reduces the warming of both the surface and boundary layer. With this reduction in heating, the temperature range on a cloud day can be substantially less than that of a clear day.

At night the cloud layer acts as a blanket by reducing the long wave radiation loss to space and increasing the long wave gain from atmospere to earth. This again reduces the temperature range compared to a clear sky night.

Surface Characteristics

The main impact that we want to look at here is snow versus no-snow situations. The albedo table above showed that snow is a good reflector. Similarly, it is a good emitter of long wave radiation with fresh snow cover being the better than old snow cover. The implication here is that the coldest nighttime temperatures will typically occur when there is snow on the ground, particularly fresh snow.

There is an old rule of thumb in Northeastern Kansas
that nighttime lows will not drop significantly below
zero (Fahrenheit) unless there is snow on the ground.
This rule assumes that there is a polar or arctic air
mass in place.

Related to the state of the earth's surface is the urban heat island effect. Studies have shown that the urban core of a city typically does not cool off as much as surrounding rural areas due to the heat retention characteristics of the urban environment (cement, etc.) Although this impact is relatively small, typically a few degrees, if you are forecasting for a large metropolitan area, this detail may be important.

Turbulent Wind Flow

In this section we want to look at the impact of turbulent wind flow on nighttime temperatures. At night, the earth's surface cools off and the adjacent boundary layer air cools via conduction. This cooling often results in a surface inversion that reflects the cooling. The air above the surface layer stays warm while the near surface air cools.

Consider what happens when the winds are fairly brisk and there is considerable mixing from the surface to several thousand feet into the atmosphere. Even though the cooling may continue at the earth's surface, the mixing brings down warmer air from above and tends to even-out and counteract the surface cooling. The net effect is that the sruface temperature does not drop as much as it might drop without the winds.

This type of situation often occurs behind cold fronts. The front ushers in colder air (cold air advection) but the winds remain fairly brisk. Until the winds weaken, the temperature drop may not be as much as expected due to vertical mixing. This effect will also depend upon the depth and strength of the post-frontal cold air.

Temperature Advection

Winds move warmer and colder air over the landscape. We measure the local change in temperature due to this movement by determining the temperature advection. If cold air advection is present in the boundary layer, temperatures are expected to drop. If warm air advection is present in the boundary layer, temperatures are expedcted to rise.

In most situations the magnitude of the boundary layer temperature advection is relatively small and the impact of other factors is stronger than temperature advection. However, there will be situations where brisk winds will make temperture advection the primary process. Without making some quantitative estimates, it is difficult to judge the relative effect of advection compared to other factors.

Latent Heat Impacts

Consider what happens when solar energy impacts wet ground. Part of that energy is used to evaporate the water in the soil instead of heating the ground and subsequently the air. The net result is boundary layer warming that is less than expected and an increase in boundary layer moisture due to the evaporation.

Another situation where latent heat impacts temperature is when warm air advection moves above-freezing air over snow cover. The snow cover by itself reduces the solar heating due to its albedo. When the above-freezing air is moved over the snow cover, the snow tends to melt and vaporize. The latent heat needed to accomplish this phase transition is taken from the air and opposes the temperature increase due to the warm air advection. Here again, temperature increases may be moderated by latent heat.


This impact could have been discussed in the previous latent heat category but is considered separately due to its more frequent occurrence.

When rain begins it often falls into and through a relatively dry layer. This results in evaporation. The latent heat for the evaporation is taken from the air and the layer tends to cool to its wet-bulb temperature. Thus, if you expect precipitation to begin, you need to assess whether evaporative cooling will result and what impact it will have on your temperature forecast.

A specific situation where this effect can be critical to your forecast is when the temperature in the low to middle 30s and rain is expected. If the wet-bulb temperature of a layer is below 32 degrees Fahrenheit, it is possible to change rain to snow because of evaporative cooling. The result is not only a temperature forecast impact, but a precipitation type impact.

Non-Diurnal Temperature Affects

Although the normal diurnal trend of a sunrise minimum and mid-afternoon maximum in temperature is the rule over 90 percent of the time, there are days when non-diurnal temperature trends occur. It is important to recognize these non-diurnal trends and incorporate them into your forecast. This section focuses on some of the factors that cause non-diurnal temperature trends.

Local Effects

Although the term "local effects" may not be the best way to describe these phenomena, things like sea or lake breezes, downslope winds, or thunderstorm downdrafts can modify the normal diurnal trends in fairly localized areas.

When a sea breeze kicks in, the winds shift and the temperature drops. This frequently happens earlier than the normal diurnal maximum temperature time. Similarly, if your forecast area is prone to downslope winds, you may get non-diurnal rises in temperature due to adiabatic warming.

Some of these effects are very localized and timing if often difficult to anticipate. Nevertheless, if a potential for these effects exists, they can be mentioned in the forecast.

Frontal Passsage

The passage of a strong front can significantly change the temperature. A strong cold front can drop the temperature by 10 to 20 degrees Fahrenheit in one hour. Similarly the passage of a strong warm front is accompanied by rising temperatures. These changes can modify the diurnal temperature trend significantly.

Strong Thermal Advection

Strong temperature advection other than that associated with fronts occur in some parts of the country. For example, it is not uncommon to have moderate warm air advection on the back side of a cold high pressure center. The rapid movement of the high to the east results in a surge of warmer air, not necessarily associated with a warm front, from the west.

One specific instance of this type of situation occurs over eastern Kansas. Perhaps once a winter, strong westerly winds occur overnight. In addition to the warm air advection, there is a downslope component that adds adiabatic warming. The net result is cooling temperatures during the evening with a gradual warming starting between 10 pm and midnight. You wake up to warmer temperatures than you went to bed with (assuming you are not working midnight shifts).

Temperature Forecasting

Most temperature forecasts these days depend on Model Output Statistics (MOS). Nevertheless there are observational data which can be used to fine-tune and modify MOS forecasts for the first forecast period. Some of these factors are discussed below.

Filling in the Morning Inversion

schematic of morning inversionThe diagram at the left represents a vertical temperature profile. The vertical axis is height increasing upward while the horizontal axis is temperature increasing toward the right.

The blue line represents a typical radiation inversion which occurs when the long wave radiational cooling is strong, the winds are light, and there is little cloud cover. The surface temperature drops off significantly compared to temperatures a few thousand feet up. The red line is a nocturnal inversion where the nighttime cooling is weaker.

The question being posed here is: Which of these two situations will exhibit a faster warming rate as solar energy impinges on the earth's surface?

Recall from discussions of thermodynamic diagrams that area is proportional to energy on some types of diagrams. This means that as heat is transferred from the earth's surface to the air above and mixed verically, the sounding will fill in as a series of temperature curves along dry adiabats. If we were to run a temperature curve from the top of the inversion along a dry adiabat to the surface, it would take more energy to warm the radiational (blue) inversion than the nocturnal (red) inversion.

On the other hand, if you apply the same amount of energy to each sounding, you would get a stronger surface temperature rise under the radiational (blue) inversion than for the nocturnal (red) inversion due to the slope of the inversion.

If you need to forecast temperatures at specific times during the day, you can use a nearby morning sounding to determine a relative rise rate.

Using the 850 mb Temperature

A technique that has been used for years to estimate the maximum temperature for the day is to start with the 850 mb temperature on a nearby 1200 UTC sounding. From that temperature follow the dry adiabat downward to the surface pressure. The temperature at that point is your maximum temperature. This technique does not take into account variations of cloud cover or the other factors mentioned above. Use it cautiously.

Model Output Statistics (MOS)

You should realize by now that a temperature forecast is intimately related to the other elements of your forecast, specifically the cloud cover, winds, and any precipitation. As you assess the first period forecast, compare the MOS forecast for all weather elements with your assessment. If there are significant differences, a modification to the MOS temperature forecast may be in order.

For example, if MOS is calling for overcast skies and moderate winds tonight, but it looks like the low pressure system is moving east faster than expected with the result that skies will clear and winds will weaken, a downward modification to the MOS low minimum temperature may be needed.

Review Questions

Instructions: Place the cursor over the answer of your choice. If you are correct, it will be highlighted in green; if you are incorrect, it will be highlighted in red.

Given the four primary factors that affect temperature, what conditions favor the coldest nights?

  1.  The air is dry.

  2.  The winds are calm.

  3.  There is fresh snow on the ground.

  4.  All of the above

Dry air allows the long wave radiation from the earth's surface
to escape into space with minimal re-emission back to the earth
by the atmosphere.

Calm wind reduces any vertical mixing and temperature advection
allowing strong surface cooling.

Fresh snow enhances the long wave radiation loss from the earth's
surface. Some of the coldest nights occur over newly fallen snow.

You live in Topeka, Kansas. Winds are expected become westerly and increase in speed by midnight. A cold high pressure center is over Illinois and moving east. How will these factors affect overnight temperatures?

  1.  Minimal impact, expect a normal diurnal fall throughout the night.

  2.  Temperatures will slowly warm after midnight.

  3.  Precipitation will counteract any warming.

  4.  Global warming has set in. Prepare for a warm winter.

Modifications to MOS temperature forecasts should take into considertion which of the following factors?

  1.  Cloud cover differences between the MOS forecast and observed data.

  2.  Wind speed differences between the MOS forecast and observed data.

  3.  Speed changes in system movement.

  4.  All of the above

Return to the Operational Weather Topics page

last updated on 3/04/10