Infrared technique: electro-optic measurements of the atmosphere (project OPAQUE) (1973 – 1981)

Electro-optic measurements on the atmosphere (project OPAQUE)

Project objectives and execution

At the end of the seventies and early eighties of the previous century, the OPtical and Atmospheric QUantities in Europe project was part of the Dutch contribution to the NATO OPAQUE project. Electro-optical measurements were performed on the atmosphere. The OPAQUE program implied that for the first time an effective statistical study was performed with corresponding measurements of the observation in both the visible and the infrared domains. While a large number of studies had already been carried out on electro-optical (EO) quantities under different meteorological conditions, the OPAQUE project was the first comprehensive collection of these quantities with corresponding observations. The measurement period for OPAQUE applied for at least two years. A number of OPAQUE stations, including the Dutch, managed to carry out these observations for a longer period of time. The OPAQUE program was organised by the members of the Defence Research Group of the North Atlantic Council NATO. A series of Research Study Groups (RSGs) of that committee developed the joint measurement program for the European area. The countries participating in the project were Canada, Denmark, Federal Republic of Germany, France, Italy, the Netherlands, the United Kingdom and the United States of America. RSG-3 of Panel III, Sky and Terrain Radiation, postponed the original start date of November 1973 because of the too short preparation time for the OPAQUE program.

OPAQUE measurement stations
OPAQUE measurement stations
  • DK, CA: a monitoring station was set up by Denmark and Canada on the Danish island Lolland. It was only a few hundred meters from the coast.
  • US, GE: measuring station Meppen, Germany was operated by the US and West Germany.
  • NE: On the former Ypenburg military airfield, this station was designed as an “urban environment” measuring station.
  • GE: This measurement site was set up by the German participants at a height of 750 m in southern Germany on the Swabian Alb south of Tübingen.
  • FR: Bruz station was set up in Brittany on the site of the CELAR (Center d’Electronique de l’Armement).
  • UK: Christchurch station was located on the English south coast between Christchurch and Bournemouth.
  • IT: The Trapani station was located at a remote location at the Birgi airport near Trapani (Sicily), just a few hundred meters from the coast with a generally windward direction.

The results of the use of electro-optical sensing and directing devices (brightness amplifiers, thermal infrared and laser systems) are influenced, among other things, by the properties of the atmosphere and the environment. To be able to numerically determine this influence, the measurement program OPAQUE was set up and executed. Seven measuring points were distributed across Western Europe (see figure) where the relevant parameters of the atmosphere and the environment were measured. These measurements were performed at every hour for a duration of four minutes both during the day and at night at corresponding sun times.

Aerial photo of the (former) airport Ypenburg (in 2003)
Aerial photo of the (former) airport Ypenburg (in 2003)

A minimum measurement program had been drawn up that had to be measured at each measurement location. In addition to specific optical parameters (transmission, brightness levels, scattering properties, etc.), the meteorological parameters were measured to study the relationship with the weather data. Because the measurement locations were spread geographically across Western Europe, the dependency of the measured parameters of climatic factors could be determined. The results of the measurements were recorded in a prescribed format on a magnetic tape. They were forwarded to a central database that was located in Great Britain. From this database, extracted data sets on magnetic tapes were distributed to the participating countries which included the measured and registered data from all measuring stations. In addition to these prescribed minimum set of measurements, additional measurements were carried out at most of the locations including the Dutch set-up. As a result, the measurement programs could vary from carrying out the standard measurements to those of performing sorties with instrumented aircraft. The NATO publication OPAQUE D-7302 (Fenn, 1987) describes the quantities to be measured, the measurement locations, the measurement plans and the data format.

A number of the measured parameters, in particular those relating to visual conditions, were also used by TNO for the FLAT project (Photochemical Air Pollution, Aerosols and Toxicity). In this project, commissioned by the then Ministry of Public Health and Environmental Hygiene at the former Institute for Environmental Hygiene and Health Technology TNO, the atmospheric conditions during the at that time very common smog formation were measured.

In December 1977, after a preparation period of several months, it was finally possible to start with the official measurement period. A minimum continuous measurement period of two years was agreed. However, the measurement program for Christchurch (England) had to be discontinued at the end of February 1979 as the laboratory moved. The measurements in Birkhof (southern Germany) were discontinued on Friday, July 13, 1979, after a lightning strike caused a lot of damage. In Meppen, the joint station of the US and the Federal Republic of Germany, measurements were carried out until the end of 1980. In Lolland, the complete two-year measurement program ran until 1 December 1979. After that, a reduced program was carried out for another year for one week per month. On 1 December 1978, the measurements at Trapini (Italy) started. Those were continued until March 1, 1981. The French and Dutch stations also remained active until that date. Much equipment of the measuring stations, which were dismantled after the executed program, was re-used at measuring set-ups in the immediate vicinity of the laboratories involved in the measurements. This applied in particular to the infrared transmission metres and the aerosol measurement equipment.

The variables measured by TNO

Variable Instrument
(name at Ypenburg)
Unit Accuracy
Visible scattering
(coefficient)
Nepheliometer AEG Point-visibility (PTHF) per km ± 20%
Transmission of the atmosphere
(fotopic 0.4 – 0.74 microns)
Transmissometer Eltro (ELTR) per km ± 10%
Infrared transfer
(3.4 – 5.0 microns)
Transmissometer Barnes (IRSC) % ± 2%
Infrared transfer
(8.0 – 12.0 microns)
Transmissometer Barnes(IRSC) % ± 2%
Cloud cover SCORPIO meter %
Temperature Surroundings (THER) C ± 0.2 C
Relative humidity Hygrometer (HYGR) % ± 5%
Dew point (humidity) German instrument C ± 1 C
Wind speed at 10 m and 2 m Dish anemometer m/s ± 2%
Rain density (RNR1 en RNR2) mm/hour
Amount of rain (PREC) mm/hour
Raindrop size Drop meter (RDRP) mm2

Photopic is a term from ophthalmology and sensory physiology. Photopic vision is the vision under good lighting circumstances. The eye pupil is narrow and the maximum light sensitivity of the retina lies at a wavelength of 555 nm (green). Colors can be observed well.

Light sensitivity curve of the human eye: blue = scotopic view (at night), red = photopic view (daytime)
Light sensitivity curve of the human eye: blue = scotopic view (at night), red = photopic view (daytime)

The transmission of the atmosphere in the visible (photopic) wavelength range

This parameter was measured with the following instruments:

  • A scatter meter. In the visible area, the attenuation of the atmosphere is usually equal to the scattering. This was measured at Ypenburg over approximate 1 metre with the aid of an AEG-Streulichtmesser (PTHF).
  • A transmissometer: an ELTRO transmissometer (ELTR) that was modified by the Royal Military Academy (KMA) with an effective measuring base of 1000 metres between transmitter and receiver. To achieve this effective measuring distance, a retro-reflector was set up 500 metres away from the combined transceiver.
  • A laser scintillometer (LASCIN) consisting of a HeNe laser and a detector (TRBM) placed at a height of 2 metres above ground level and standing at a distance of 100 metres (see photo below). This commercial laser was a 2 mW laser with an opening angle of 1 mm2 and a beam divergence of approximately 1.6 mRad. The detector had an effective area of 5.1 mm2 and the opening angle was 12.5 mRad. To eliminate the error caused by backlighting, a 633 nm filter was placed in front of the detector (the so-called Rinkema box).
Rinkema-box
Rinkema-box

 

 

 

 

 

 

 

Part of the OPAQUE setup at Ypenburg
Part of the OPAQUE setup at Ypenburg

Four different rain gauges are visible in the foreground. In the building behind it are the visual and infrared transmissometers. Above it the receiving antenna of the transmissometer for mm-radar waves.
 

The transmission of the atmosphere in the infrared wavelength range

With a transmissometer from Barnes, Model 14-708 (IRSC), the weakening of the atmosphere in the infrared wavelength with a narrow band of 4.0-4.1 μm and the bands 3.4 – 5.0 μm, 8.0 – 12.0 μm and 8.25 – 13.2 μm was determined. A black heat source (temp. + 650 0C) was used as an infrared source (IR). The receiver was a non-cooled thermistor detector at 500 metres distance from the heat source. The Model 14-708 contained all components that were required for accurate calibration. To calibrate, it was only necessary to move the IR source close to the receiver and align them optically. Then a built-in diaphragm was adjusted causing the receiver to see the source as if it were 1 km away. The receiver was then set so that the digital display reproduced 100% of the radiation emitted by the source in the four spectral bands. Measuring the radiation in each of the four spectra took about a minute; four minutes for a complete measure.
This instrument is discussed separately in combination with the measurements in the visible spectrum and the (radar) millimetre wave area (94 GHz).

Image of the infrared transmission measuring instrument
Image of the infrared transmission measuring instrument

The illuminance

Also in the visible wavelength range, the illuminance was measured in a horizontal plane and in four vertical planes placed in the wind directions (N, E, S and W). At Ypenburg, a Physics Laboratory RVO-TNO and the Royal Military Academy instrument (LUXM) was installed. It connected a lux-meter via a clear plastic conductor and a rotating light coupling to a set of measuring windows in the aforementioned five directions. Twelve additional directions (see figure and photo below) were provided. (see figure and photo below).

The LUXM schematic
The LUXM schematic

The LUXM
The LUXM

 

The scattering of the atmosphere (luminance) during the day (PVIS)

The amount of radiation that was scattered in the photopic wavelength range (0.34-0.74) in the four wind directions (N, E, S, W) due to incident light (sun, sky) was measured. A measuring instrument (PVIS) was developed by the University of California (San Diego) for this purpose and was installed at Ypenburg. A sensitive optical receiver equipped with a photomultiplier ‘looked’ in a room with a very low reflection coefficient that was placed at a distance of 23 centimetres. The entire device was mounted rotatable about a vertical axis (see figure).

Luminance meter
The principle of measuring the atmospheric luminance

 

The scattering of the atmosphere (luminance) at night (NPRD)

The same sensitive optical receiver mentioned above “looked” over a distance of 100 metres in an easterly direction in a space with a very low reflection coefficient. With a TeleLuminanceMeter developed by the Physics Laboratory RVO-TNO, the measured luminance of the optical receiver was compared with that of a built-in stabilised weak light source. By applying a kind of compensation with a grey wheel and a number of filters driven by a servomotor, it was possible to realise a dynamic range of ten decades. This allowed an overlap with the scattering measured in the daytime only in the eastern direction.

The overall transmission through an atmospheric column

Solar radiation was measured in a number of wavelength ranges (PYRH). In addition, measurements were made in the open and closed state of the receiver. The meter was mounted on an astronomical table and remained focused on the sun. In the night, the table was reversed 3600. Eppley pyrheliometers of the thermoelectric type were placed at all OPAQUE measuring stations for these measurements. From the response at the various wavelengths, it was expected that conclusions could be drawn about the composition of the atmosphere. The radiation was collected on two concentric silver rings, the outer surface of was covered with magnesium oxide (white surface) and the inner ring with lamp black. A construction with thermocouples (thermopile) was used to measure the difference in temperature between the two rings. Provisions were also made to measure the direct and diffuse solar radiation.

Pyrheliometer: photopic 0,55 µm, 0,87 µm, 0,945 µm, 1,06 µm, 0,40 µm and 0,75 µm.
Pyrheliometer: photopic 0,55 µm, 0,87 µm, 0,945 µm, 1,06 µm, 0,40 µm and 0,75 µm.

The meteorological parameters

To determine the relationship between the aforementioned parameters and the weather, the following meta-parameters were measured at all OPAQUE stations at every hour during the day:

  • temperature
  • atmospheric pressure
  • relative humidity (Hygrometer)
  • dew point temperature
  • wind speed and wind direction at heights of both 2 and 10 meters
  • rainfall (mm/hour)
  • rain speed (integration time ~ 1 min)

In addition, the cloud level and the ground state were determined if possible. The cloud formation strength or the percentage degree of coverage was measured with the “Cloud scanner SCORPIO“.

Thermometer
Thermomometer

Hygrometer
Hygrometer

 

The composition of particles in the air

The Atmospheric Sciences Laboratory, White Sands Missile Range, New Mexico, installed a so-called Atmospheric Particulate Collector at all OPAQUE stations, with which air particles were sucked in and deposited on a filter. These filters were replaced every week and sent to White Sands for analysis.

Additional measurements

Additional measurement programs were carried out at many OPAQUE stations. The measuring frequency was often chosen to be larger than the mandatory once per hour (e.g. the transmission measurements during low visibility periods) or the prescribed parameter was measured in several directions. An aerosol counter was placed at virtually every measurement location. The size distribution of the aerosols (particles in the air) was measured with an arrangement composed of a measuring cabin with a suction device and a Model 225 Royco Optical Particle Counter 10. These particles have a large influence on the transmission and scattering properties in both the visible and the infrared wave domains.
The Air Force Geophysics Laboratory, Bedford, Massachusetts in the US, conducted flights with an instrumented Hercules C-130 aircraft. With this aircraft, optical and atmospheric parameters were measured as a function of height in the vicinity of the OPAQUE measuring stations.

Het DFVLR (Deutsche Forschungs- und Versuchsanstalt für Luft und Raumfahrt in Oberpfaffenhofen) carried out measuring flights daily during a specific week. These flights were made between South and North Germany via Rotterdam, among other things, to measure the lateral variations and correlations between the parameters.

At the German measurement site at Birkhof and at Ypenburg the transmission properties of the atmosphere for the radar mm-wave area were measured. To this end, raindrop counters were also set up, with which the size distribution of the raindrops was determined. In addition, cloud cameras and equipment to determine the CO2 concentration were established at a number of OPAQUE measurement locations.

Processing and analysis of the measured data

After the measurement program got off to a good start, the work shifted more towards the processing and analysis of the measurement data. From September 1978 a special group of people worked on data processing. The investigations were specifically focused on:

  • the determination of the statistics of the measured parameters,
  • the determination of any links between the measured parameters,
    in particular those between the optical parameters and the meteorological data,
  • verifying existing models for transmission through the atmosphere, such as the LOWTRAN model (Low-Resolution Atmospheric Radiance and Transmittance),
  • the preparation of improved models for transmission through the atmosphere,
  • the preparation of prediction methods for the optical parameters of the atmosphere from the weather forecast,
  • drafting models for the application of electro-optical sensors under different climatic conditions.

The first measurement results

At the end of January 1980, it turned out that the measurement results were very satisfactory. There had been a small number of failures. The Ypenburg station reached about 90% of the possible measurements. The station worked fully automatically with the help of an Alpha CA minicomputer. An inspection took place once per working day and the cassette tape with the hourly measurement data was retrieved for further processing.
To increase the accuracy of the measurements as high as possible, the results were compared with those of mobile equipment. Special attention was paid to the calibration of the infrared transmissometer. At the airport Pershore (Malvern, UK) all Barnes transmissometers were set up and compared in parallel. The spectral sensitivity of the receivers was also measured. Through these so-called “Intercomparison Trials” the mutual comparison of the measurement results was greatly improved.
A problem with the calibration of the 100% point of the infrared transmissometer could not be solved. The provision made by the manufacturer turned out to be insufficient. A solution was found by equating the favourable conditions (good visibility, low relative humidity) with the value calculated with the LOWTRAN program. It was estimated that the uncertainty in the accuracy for the infrared transmission achieved was approximately 3%.

The transmission of the atmosphere in the infrared

because virtually no statistical data were available about the transmission through the atmosphere in the infrared wavelength range, these data were developed first. The figures below show firstly a histogram of the time that the infrared transmission in the 3.4-5.0 μm
range is between the values specified along with the abscissa. Second, the cumulative distribution function of the transmission in the 3.4-5.0 μm range is shown.

Histogram of the transmission in the 3,4-5,0 μm area during September 1978 (distance 500 metres)
Histogram of the transmission in the 3,4-5,0 μm area during September 1978 (distance 500 metres)

Cumulative distribution function of the transmission in the 3,4-5,0 μm area during September 1978 (distance 500 metres)
The cumulative distribution function of the transmission in the 3,4-5,0 μm area during September 1978 (distance 500 metres)

Histogram of the transmission in the 3,4-5,0 μm area during November 1978 (distance 500 metres)
Histogram of the transmission in the 3,4-5,0 μm area during November 1978 (distance 500 metres)

Cumulative distribution function of the transmission in the 3,4-5,0 μm area during November 1978 (distance 500 metres)
The cumulative distribution function of the transmission in the 3,4-5,0 μm area during November 1978 (distance 500 metres)

 

Overview of the transmission over 500 metres in the 8.0 to 12.0 μm range from March 1977 to February 1978. The vertical axis indicates the time fraction in which the transmission was lower than the parameter value.
Overview of the transmission over 500 metres in the 8.0 to 12.0 μm range from March 1977 to February 1978. The vertical axis indicates the time fraction in which the transmission was lower than the parameter value.

The histograms above show that in the month of November of 1978 the infrared transmission was clearly worse than in September 1978. November was a month in which there was a lot of fog. Although the infrared transmission is better during fog and thin fog than for visible radiation, there is a clearly negative influence. Similar histograms have been made for the 8-12 μm transmission window. The transmission is generally more favourable than in the 3.4-5.0 μm range. Also in this wave area, there is a clear reduction for the transmission in fog. In the figure below, in the vertical direction, the fraction of time is given, that the infrared transmission is lower than the value that applies to the drawn curve. The month of September 1978 is missing. All infrared transmissometers were then compared in England. From the overview, it can be read that in November 1978 the chance was 0.15 that the transmission was less than 50% (over 500 metres distance).

Comparison between the transmission in the infrared and visible wave area

Scatterplot of the infrared transmission between 3.4-5.0 μm against the visual transmission, measured over 500 metres distance
Scatterplot of the infrared transmission between 3.4-5.0 μm against the visual transmission, measured over 500 metres distance

In the accompanying figure, the transmission in the 3,4-5,0 μm area is plotted along the vertical axis against (in the horizontal direction) the transmission in the visible area, as measured with the Eltro-transmissometer. From this so-called “scatter plots”, relationships between the plotted quantities can be determined. As can be seen, there is no direct relationship between the two quantities. In general, the transmission in the visible area is lower or at most equal to that in the infrared range. For low transmittance values in the visible range, the infrared transmission can vary greatly between high and low values. The size of water droplets (aerosols) is very likely to play a major role in this.

The relationship between the infrared transmission and the humidity of the atmosphere

In the first figure below, in the vertical direction, the infrared transmission in the 8.0 to 12.0 μm range is plotted against the absolute humidity along the horizontal axis.

"Scatterplot

 

Scatter plot of the transmission over 500 metres distance between 8.0 to 12.0 μm against the relative humidity
Scatter plot of the transmission over 500 metres distance between 8.0 to 12.0 μm against the relative humidity

In the literature, it was generally assumed that the transmission (both in the 4 μm and in the 10 μm window) depends on the total amount of water vapour present in a column between transmitter and receiver. There is also a decreasing tendency in the transmission observable with increasing water vapour density. Remarkable here is a large number of outliers down along the vertical axis. Later it was checked whether that was caused by rain or fog.
The relationship becomes much more obvious when the absolute humidity is not plotted along the horizontal axis, but the relative humidity is plotted. The infrared transmission is now almost constant for values of the relative humidity ≤ 95%. The infrared transmission can be used for values >95% between 0 and 1. An explanation for this may be that at large relative humidity, where fog often occurs, the size of the mist droplets strongly influences the infrared transmission. As soon as these droplets have the same size as those of the wavelength of the infrared radiation, a strong absorption and scattering occur.

Light levels

In order to be of service to the analysis of the OPAQUE data, a computer program was developed at the then Physics Laboratory RVO-TNO. This program had as input parameters the date, the time and the local geographic coordinates. As an output, this program, called ILLUM provided:

  • the azimuth of the sun,
  • height (altitude) of the sun,
  • the azimuth of the moon,
  • altitude (altitude) of the moon,
  • the phase of the moon,
  • the illuminance on a horizontal plane by sunshine in clear weather,
  • the illuminance on a horizontal plane by the moonshine in clear weather,
  • a warning about an eclipse.

More measurement results

More measurement results can be found in “Comparison of simultaneous atmospheric attenuation measurements at visible light, mid-infra-red (3-5 µm) and millimetre waves (94 GHz)“.

In the analysis of the data collected by each country in the US, the following was put forward:

The Ypenburg airfield, Netherlands, and Christchurch, England, OPAQUE were selected for analysis because of the number and completeness of available data. Ypenburg produced 47 months of data (from March 1977 to February 1981, excluding June 1980). Christchurch produced 27 months of data (from December 1976 to February 1979). Ypenburg (52 0 03 ‘ N, 4 0 22’ E ) is located about 7 km away from the centre of The Hague and 10 to 15 km northwest of Rotterdam. The site is strongly influenced by artificial illumination at night. The Physics Research Group of the TNO Physics Laboratory managed the OPAQUE measurement program at the Dutch site. Ypenburg is representative for an urban industrial environment in northern Europe. The UK OPAQUE site located at the Royal Signals Research Establishment (RSRE) at Barnsfield Health, near Christchurch, was located about 8 km inland from the south coast of England (5 0 41 ‘ N, 1 0 45 W. Christchurch is representative or northern Europe’s maritime environment.

 

 


Background information: contrast and angle size

An important point in collecting the data is that the threshold for daylight contrast changes very little for the human eye as a function of the backlight, but is highly dependent on the angle size (size of the object), including an object against this background is seen and also depends on the time with which one looks at the object.

Angle size as a function of the contrast for the probable view of 99% in the lack of knowledge about the position of the object by ± 4 degrees or more (Gordon 1979)
Angle size as a function of the contrast for the probable view of 99% in the lack of knowledge about the position of the object by ± 4 degrees or more (Gordon 1979)
Comparison of transmissometer measurements with human observations during the day e0 = Cr (Hering et al., 1971)
Comparison of transmissometer measurements with human observations during the day e0 = Cr (Hering et al., 1971)

A well-chosen assumption for both psychoanalytic and physical measurements is that the view with the probability bordering on certainty can be defined as the distance to a black object with reasonable dimensions (contrast = -1), on which this object can be seen against a cloudless
horizon. This is shown in relation to the attenuation α with an accuracy of +/- 20% in the equation below. Is there a reasonably constant threshold in the contrast | C r | then the graph will show a descending straight line on a logarithmic scale. Displayed with the following formula: -ln |Cr| = ßor simplified for a threshold contrast of 0.05 (=angle of 0,5 degree): ßr = 3.0 +/- 0.