7.3. How-To
7.3.1. THERM
The following two sections were previously reported in [Schott 1993].
The fundamental purpose of the thermal sub model is to calculate the time dependent temperatures of objects within the scene as influenced by their environment. THERM is a linear differential temperature generation model written by [DCS 1990] tested for validity at RIT [Spector 1991]. Temperatures are calculated separately for each facet as a function of time based on first principles models which determine the rate of heat transfer corresponding to a specific temperature difference between an object and its environment. Each facet is assumed to be thermally independent of the other's and exhibit an isothermal surface behavior. THERM predicts an object's thermal signature based on solar parameters, meteorological conditions, and material properties. As a stand alone model, THERM has been shown to produce accurate temporal predictions of temperatures of real world objects.
THERM requires various object and environmental parameters for input to the linear differential heat transfer equation in order to calculate the effects of radiative heat exchange, convection, conduction (within the facet only), and absorption of visible insolation (solar radiation). Each parameter impacts in some way the temporal temperature of the various objects within the scene.
7.3.1.1. Environmental Parameters
The environmental parameters can be divided into two categories, the location parameters and the meteorological parameters. The location parameters include (1) latitude & longitude, (2) date, time, and time interval, and (3) time of sunrise / sunset. THERM will calculate the sunrise / sunset times from the latitude, longitude, date, and time parameters if necessary. The meteorological parameters include (1) direct & diffuse insolation along with high noon transmission, (2) air temperature (sunrise, peak, and peak time), (3) air pressure, (4) relative humidity, (5) dew point, (6) wind speed, (7) sky exposure (calculated as the percent of sky not obscured by clouds), (8) cloud type (as either cirrus, cirrostratus, altocumulus, altostratus, stratocumulus, stratus, and fog), and (9) precipitation type / rate / temperature.
7.3.1.2. Object Parameters
The object parameters consist of (1) heat capacity, thermal conductivity, and thickness to determine thermal mass (which impacts the rate at which the temperature of an object can react to a given amount of heat), (2) self-generated power, (3) exposed area and slope & azimuthal angles, and (4) solar absorptivities. These parameters must be supplied for each material type contained in the scene
Table 7-1. Inputs to Thermal Submodel
| Location Parameters | Material Parameters |
|---|---|
| Latitude* | Heat Capacity* |
| Longitude* | Thermal Conductivity** |
| Date* | Thickness* |
| Time (Difference from GMT)* | Solar Absorptivities* |
| Time Interval* | Exposed Area* |
| Time of Sunrise | Self Generated Power* |
| Time of Sunset | Slope and Azimuthal Angles* |
| Meteorological Parameters |
|---|
| Direct Insolation |
| Diffuse Insolation |
| High Noon Transmission |
| Air Temperature** |
| Sunrise Air Temperature* |
| Peak Air Temperature* |
| Peak Air Temperature Time* |
| Air Pressure** |
| Relative Humidity** |
| Dew Point** |
| Wind Speed** |
| Sky Exposure** |
| Cloud Type** |
| Precipitation Type/Rate/Temperature** |
*Required input parameter ** Can be input as a temporal file or as a single value for THERM to compute temporal values which can be edited by the user
THERM can either accept the input of temporal meteorological data, i.e. a weather history, or can compute the temporal data using simple environmental models when supplied a limited input. With the first method, the location parameters are input along with a file containing the remaining temporal meteorological data. Figure 7-3 shows the temporal nature (diurnal cycle) of various meteorological parameters and the effects that one parameter can have on another. Inputting temporal data is the most accurate method as it provides the sub model with the maximum amount of information needed to create a weather history and predict temporal object temperatures weather history. However, such data is not always available, particularly when one is trying to simulate a future day's scenario. In such cases THERM will predict the necessary temporal parameters given a limited input of average parameter values. Usually, an average value over the entire day for the parameter in question is selected unless a specific value is requested, such as the maximum temperature for the day. Given the inputs for one point in time, the program computes an estimate of the weather history for the entire day. THERM can then calculate the direct and diffuse insolation from these estimated temporal environmental and location parameters including latitude, season, sun elevation, cloud type, sky cover, air temp, air pressure and humidity. From these data, THERM can create an estimated weather history and predict temporal object temperatures. This estimated weather history can be further refined by the user to correspond to a specific scenario's requirements, such as rain, cloud cover, etc., thus allowing the analyst to observe the effects of such variables on the final image.
The weather history for a 24 hour period or longer is initialized using one of the two methods described above. THERM proceeds to compute the temperature of object facets at any point in time based upon the input parameters, either estimated or actual. Since there are no "real" initial object temperatures when THERM starts the process, an approximation is made by setting object temperatures to their initial equilibrium value. For purposes of accurate temperature prediction, it will be necessary to start the extrapolation process several hours before the desired time of simulation. This will allow the model to reach a dynamic equilibrium where the effects of all the parameters in THERM interact to reach a stable contribution to the object's temperature