Use of Special Purpose Infrared Thermometer to Measure Downward Longwave Radiation from Greenhouse Gases

Background

A key factor in future estimates of increased global temperatures resulting from higher levels of atmospheric carbon dioxide is its relationship with water vapor and clouds in the absorption of heat flux leaving the Earth’s surface and the resulting radiation back to the surface. In particular, the relative importance of clouds versus all of the greenhouse gases has been difficult to verify. Recently, the instrumentation industry has developed low cost infrared thermometers that are finding wide use in many industrial applications involving non contact temperature measurements. These devices measure incoming infrared radiation energy (watts per square meter) over a limited bandwidth, and using Planck’s equation and special fitting equations in the electronics, convert this to the target’s temperature.

The purpose of this document is a recommendation for the development of a similar special infrared thermometer that will enable the measurement of heat flux from the atmosphere under a variety of conditions and thereby evaluate various heat balance climate models and possibly lead to improved models in the future. Preliminary results indicate some models substantially overestimate downward radiation from clear skies where greenhouse gases predominate.

Results of Preliminary Work

A standard commercial IR thermometer with a range down to -76F (the lowest now available) was used to measure the “apparent” temperature of the atmosphere from the Earth’s surface under a limited variety of surface temperatures and cloud conditions. The first thing that was noticed was a large difference between the case of clouds or part clouds and clear skies. Low thick clouds can have a temperature of only about 5F to 15F below the surface temperature. Apparently with an emissivity near one, these low cloud temperatures are real and mostly indicate the altitude of the clouds implied from their temperature. The interesting cases are those with clear skies because this is when the role of greenhouse gases can be evaluated without the overriding influence of clouds. It was noted that as little as 20% cloud cover would raise the measured temperature substantially. Fully clear skies don’t occur very often in many locations. In Minnesota recently in winter, many days would elapse before one without clouds. When a clear day occurred, the measured temperature was below the measurement range of the instrument (-76F).

Later in Arizona with clear skies and surface temperatures in the 60 to 70F range, clear skies read from -60 to -70F. As explained in the next section next, these readings have to be corrected for the emissivity of the atmosphere over: (1) the IR spectrum of the sensor and (2) the full spectrum of the atmosphere at its involved temperature. It is interesting to note that after these readings were converted to heat flux being radiated down, the results were approximately 178 watts per square meter. This compares to that of 278 watts per square meter estimated in an excellent paper by Kiehl and Trenberth,¹ from a clear sky with a similar surface temperature of about 59F. In the same paper, with cloudy skies a value of 324 watts per square meter was estimated. This corresponds to a cloud temperature of about 25F cooler than the surface, a value well within the ranges of that observed with cloudy skies. The significance of the reduced down radiation compared to that expected from a clear sky indicates that greenhouse gases may have a smaller role in heat trapping relative to clouds than various models predict. Therefore increases in their concentrations will also have a reduced effect on global temperature rise compared to previous estimates.

Proposal for the Development of a Special Infrared Thermometer for Atmosphere Measurements

It is proposed that funds be supplied for cooperation between climate scientists familiar with greenhouse gas spectral absorption and companies that develop IR thermometers so that heat flux in watts per square meter from the atmosphere to the Earth’s surface can be measured directly. Some post processing to get the final result may be needed based on the surface air temperature and dew point. The second phase of this effort could include dispersal of a large number of these to many locations including in ocean areas. Further enhancements would combine this with measurements of percent cloud cover and cloud altitude at the same location and later automatic wireless transmitting of this data to a central location. Or the devices themselves could be used to estimate cloud cover by inference from the heat flux received. This could provide useful information on day/night cloud cover differential, an important factor now believed unaccounted for in earth heat balance estimates. Day time solar heat reduces cloud area which increases surface heat gain; at night clouds cool and increase in size which reduces surface heat loss. This will result in different outcomes from models that use average cloud cover for day and night. It would be expected to show more net incoming solar heat and less down radiation from greenhouse gases in order to obtain a long-term global heat balance.

Compensating for the Atmosphere Emissivity

When an IR thermometer converts the amount of heat energy it captures into a temperature reading, it is done with the assumption that the target emissivity is constant over the wavelengths that correspond to the temperature being measured. In other words the target is assumed to be either a black body or a gray body of known constant emissivity. The fact that the sensor covers only a limited bandwidth is not a concern and can be handled with appropriate equations as long as the sensor output signal is large enough but not near saturation (the temperature is in the sensor’s range). However, when measuring the atmosphere containing various greenhouse gases the emissivity will vary over the sensor’s bandwidth. This will reduce the apparent temperature, the one displayed by the IR thermometer. Compensation for this can be done by knowing how the atmosphere emissivity varies over the sensor’s bandwidth. Once the compensated temperature is known, the corresponding actual watts per meter can be estimated by knowing the emissivity of the atmosphere over the spectrum corresponding to the actual (compensated temperature).

The following equation applies:

Wa = We / em * ea

Where Wa = estimated watts per square meter coming from the atmosphere

We = the calculated watts per square meter from the indicated measured temperature

ea = the emissivity of atmosphere integrated over the sensor’s bandwidth power density

ea = the emissivity of atmosphere integrated over the measured temperature bandwidth power density

Example Measurements and Calculations

The IR meter used here has a reported sensor bandwidth of 3 to 5 microns. The detailed absorption over this range was not available, so it was assumed to be complete. The heat absorbed over the approximate atmosphere of 10 km path at 45 degrees starting a surface of 60F and 50% RH was estimated at 0.494 of the total over the 3 to 5 micron bandwidth and 0.686 over the bandwidth corresponding to the surface temperature of 3 to 70 microns. These calculations were made based on spectral absorption runs from SpectralCalc.com of GATS Corp. of the atmosphere with typical amounts of water vapor and carbon dioxide at temperatures and pressures corresponding to average values over the 10 km path. In a Phoenix Arizona location with average one day/night surface air temperature of 65F and dew point of 36F, an average day/night atmosphere clear sky readings with the IR thermometer were -66F.

We = from Stephan’s-Boltzmann at -66F or 218.7K = 128 watts/m-2

Wa = estimated actual heat flux = 128 / 0.494 * 0.686 = 178 watts/m-2

Some sensors operate over a larger bandwidth such as a 2 to 20 micron range, and these may require less correction resulting in improved final accuracy.

Over a two-week period there were 23 readings taken at various times of day with clear skies including several with a few thin high cirrus clouds. The mean value was 181 watts/m-2 with a standard deviation of 9.5 and a range of 168 to 207. Surface temperatures varied 52F to 79F and dew points from 37F to 45F. Both these readings and previous measurements show that there is a direct correlation to the surface temperature and the measured atmosphere temperature and therefore the downward radiation.

Prepared By

This document was prepared by Richard J. Petschauer, a retired electrical engineer who is interested in additional studies and measurements relative to the role of increased levels of carbon dioxide on future average global temperatures. He is not proposing to do the work recommended here but hopes others will further pursue this possible fruitful line of evaluation.

Richard J. Petschauer 7520 Cahill Road #114 Edina MN 55439 952-941-3553 rjpetsch@aol.com