The Stabilized Radiometer Platform (STRAP)—An Actively Stabilized Horizontally Level Platform for Improved Aircraft Irradiance Measurements

Anthony Bucholtz Naval Research Laboratory, Monterey, California

Search for other papers by Anthony Bucholtz in
Current site
Google Scholar
PubMed
Close
,
Robert T. Bluth Center for Interdisciplinary Remotely Piloted Aircraft Studies, Naval Postgraduate School, Monterey, California

Search for other papers by Robert T. Bluth in
Current site
Google Scholar
PubMed
Close
,
Ben Kelly L-3 Communications Sonoma EO, Santa Rosa, California

Search for other papers by Ben Kelly in
Current site
Google Scholar
PubMed
Close
,
Scott Taylor L-3 Communications Sonoma EO, Santa Rosa, California

Search for other papers by Scott Taylor in
Current site
Google Scholar
PubMed
Close
,
Keir Batson L-3 Communications Sonoma EO, Santa Rosa, California

Search for other papers by Keir Batson in
Current site
Google Scholar
PubMed
Close
,
Anthony W. Sarto L-3 Communications Sonoma EO, Santa Rosa, California

Search for other papers by Anthony W. Sarto in
Current site
Google Scholar
PubMed
Close
,
Tim P. Tooman Sandia National Laboratories, Livermore, California

Search for other papers by Tim P. Tooman in
Current site
Google Scholar
PubMed
Close
, and
Robert F. McCoy Jr. Sandia National Laboratories, Livermore, California

Search for other papers by Robert F. McCoy Jr. in
Current site
Google Scholar
PubMed
Close
Full access

Abstract

Measurements of solar and infrared irradiance by instruments rigidly mounted to an aircraft have historically been plagued by the introduction of offsets and fluctuations into the data that are solely due to the pitch and roll movements of the aircraft. The Stabilized Radiometer Platform (STRAP) was developed to address this problem. Mounted on top of an aircraft and utilizing a self-contained, coupled Inertial Navigation System–GPS, STRAP actively keeps a set of uplooking radiometers horizontally level to within ±0.02° for aircraft pitch and roll angles of up to approximately ±10°. The system update rate of 100 Hz compensates for most pitch and roll changes experienced in normal flight and in turbulence. STRAP was mounted on a Twin Otter aircraft and its performance evaluated during normal flight and during a series of flight maneuvers designed to test the accuracy, range, and robustness of the platform. The measurements from an identical pair of solar pyranometers—one mounted on STRAP and the other rigidly mounted nearby directly to the aircraft—are compared to illustrate the accuracy and capability of the new platform. Results show that STRAP can keep radiometers level within the specified pitch and roll range, that it is able to recover from flight maneuvers outside of this range, and that it greatly increases the quantity of useful radiometer data from any given flight. Of particular note, STRAP now allows accurate measurements of the downwelling solar irradiance during spiral ascents or descents of the aircraft, greatly expanding the utility of aircraft radiometer measurements.

Corresponding author address: Anthony Bucholtz, Naval Research Laboratory, 7 Grace Hopper Avenue, Monterey, CA 93943. Email: anthony.bucholtz@nrlmry.navy.mil

Abstract

Measurements of solar and infrared irradiance by instruments rigidly mounted to an aircraft have historically been plagued by the introduction of offsets and fluctuations into the data that are solely due to the pitch and roll movements of the aircraft. The Stabilized Radiometer Platform (STRAP) was developed to address this problem. Mounted on top of an aircraft and utilizing a self-contained, coupled Inertial Navigation System–GPS, STRAP actively keeps a set of uplooking radiometers horizontally level to within ±0.02° for aircraft pitch and roll angles of up to approximately ±10°. The system update rate of 100 Hz compensates for most pitch and roll changes experienced in normal flight and in turbulence. STRAP was mounted on a Twin Otter aircraft and its performance evaluated during normal flight and during a series of flight maneuvers designed to test the accuracy, range, and robustness of the platform. The measurements from an identical pair of solar pyranometers—one mounted on STRAP and the other rigidly mounted nearby directly to the aircraft—are compared to illustrate the accuracy and capability of the new platform. Results show that STRAP can keep radiometers level within the specified pitch and roll range, that it is able to recover from flight maneuvers outside of this range, and that it greatly increases the quantity of useful radiometer data from any given flight. Of particular note, STRAP now allows accurate measurements of the downwelling solar irradiance during spiral ascents or descents of the aircraft, greatly expanding the utility of aircraft radiometer measurements.

Corresponding author address: Anthony Bucholtz, Naval Research Laboratory, 7 Grace Hopper Avenue, Monterey, CA 93943. Email: anthony.bucholtz@nrlmry.navy.mil

1. Introduction

For many years radiometers that measure solar or infrared (IR) irradiance have been mounted on various research aircraft in attempts to characterize the radiative budget within the atmospheric column (e.g., Fritz 1948; Kuhn and Suomi 1958; Roach 1961; Albrecht et al. 1974; Cox and Griffith 1979; Ackerman and Cox 1981; Rawlins 1989; Saunders et al. 1992; Hayasaka et al. 1995; Valero et al. 1997; Hignett et al. 1999; Wendisch and Keil 1999; Valero et al. 2003). Typically, this has been done by mounting an identical set of broadband, or spectrally resolving, solar or IR radiometers directly to the top and bottom of an aircraft to simultaneously measure the down- and upwelling solar or IR irradiance at a given altitude. However, to properly measure irradiance requires by definition that the instrument be horizontally level, which is rarely attained on an aircraft. During turns, climbs, and descents the pitch and roll angles of an aircraft will vary. In mild to severe turbulence an aircraft will experience random changes in pitch and roll that vary in both time and magnitude. Even during so-called straight and level flight at a constant altitude an aircraft may fly in a slightly nose-up or nose-down attitude. Scientific research flights are particularly problematic since they usually involve considerable flight maneuvering (numerous turns, ascents, and descents) under a variety of weather conditions (from clear, calm skies to cloudy, convective conditions) that can cause slight to severe turbulence. Radiometers rigidly connected to an aircraft will tilt in pitch and roll along with the aircraft causing offsets and fluctuations in the data that are unrelated to any changes in the radiative state of the atmosphere but are solely due to the varying, nonlevel orientation of the instruments. The magnitude of these errors can vary greatly depending on the degree of tilt of the radiometers from a horizontally level position, the angular characteristics of the atmospheric radiation field (e.g., the height of the sun in the sky, the cloud conditions, etc.), and the angular response of the instruments. For very steep turns of the aircraft with the sun low in the sky, such as near sunrise or sunset, the measured solar irradiance can increase or decrease by as much as 100%. Typically, however, for more nominal flight maneuvers, with the sun higher in the sky, the attitude induced errors fall in the range of 5%–30%. For a measured downwelling solar irradiance of 1000 W m−2 (a typical, approximate value for clear skies, midday, at an altitude of 1000 m) this represents errors in the range of 50–300 W m−2. Such large variations are usually the dominant error term in the measured signal and one of the most significant factors in limiting the accuracy and useful yield of aircraft irradiance measurements.

Methods have been developed to try to minimize and correct for these errors. These methods include attempting to keep the aircraft as level as possible while making measurements, ignoring data when the pitch and roll angles of the aircraft are too large, and attempting to correct the data back to a level platform in postprocessing using coordinate transformations and iterative schemes (Hammer et al. 1991; Bannehr and Glover 1991; Saunders et al. 1992; Bannehr and Schwiesow 1993; Valero et al. 1997; Boers et al. 1998). However, these methods can be onerous, time-consuming, and ineffective in completely removing the errors in the measurements. When sky conditions are inhomogeneous, for example under partly cloudy skies, these methods may fail completely because of the difficulty in separating out changes in signal due to the varying attitude of the aircraft from the changes due to the varying cloud conditions.

It has long been recognized that if radiometers could be kept horizontally level in flight the accuracy of solar and IR irradiance measurements from aircraft would significantly improve and the data analysis would be greatly simplified. Luckily, a similar need for stabilization of aircraft-mounted cameras and sensors has existed for many years in fields such as aerial photography, photographic surveying, law enforcement surveillance, and television news and movie work (Austen 2002). These industries have driven advancements in airborne stabilization technology such as fiber-optic gyros and fast, accurate, magnetic torque motors that can be exploited for airborne atmospheric science research. Wendisch et al. (2001) describe an active, horizontally stabilized system they developed that keeps an airborne spectral albedometer level to better than ±0.2° for pitch and roll angles within the range of ±6°. This was the first such system developed specifically for aircraft radiometry. They employ an artificial horizon (AHZ) that consists of three linear servo-acceleration sensors and three fiber-optic gyros to give the position and attitude of the aircraft, and a global positioning system (GPS) to correct for drift in the AHZ. The information from the AHZ–GPS system is used to keep the 2D tilt stages holding the optical inlets of their albedometer level with respect to earth-centered coordinates. Wendisch et al. (2001) illustrate the improvement in the irradiance measurements when the instruments are horizontally stabilized during flight. Their leveling platform has been successfully applied to the measurement of radiation profiles (Wendisch and Mayer 2003) and the spectral surface albedo (Wendisch et al. 2004). While the design of their initial system had each 2D tilt stage holding only one sensor with a weight of less than 1 kg (M. Wendisch 2003, personal communication), an advanced and revised version simultaneously levels sensors that measure irradiance and actinic flux density (Jäkel et al. 2005).

Here we describe a newly developed, actively stabilized, horizontally level platform for making highly accurate aircraft radiometer measurements of the downwelling solar and IR irradiance. Mounted on top of an aircraft, the Stabilized Radiometer Platform (STRAP) keeps a set of uplooking radiometers level while the aircraft pitches and rolls during flight. While sharing some of the design features of the system of Wendisch et al. (2001), specifically the use of a coupled inertial navigation system (INS)–GPS to give information on the attitude of the aircraft and platform, STRAP was developed completely independently through a collaboration between the Naval Postgraduate School (NPS) Center for Interdisciplinary Remotely Piloted Aircraft Studies (CIRPAS), the Naval Research Laboratory (NRL), and the Sandia National Laboratories (SNL). It was designed and built by L-3 Communications Sonoma EO (formerly Sonoma Design Group), hereafter referred to as Sonoma EO. A prototype version of the stabilized platform was initially designed and built by Sonoma EO in collaboration with the Department of Energy’s (DOE) Atmospheric Radiation Measurement–Unmanned Aerospace Vehicle (ARM–UAV) Program at SNL. STRAP is a second-generation version of the SNL platform and was built for CIRPAS. Its design was modified to incorporate lessons learned from the development of the SNL prototype and to extend the range of angles at which active leveling of the instruments is possible. For example, STRAP has a more modular design that makes it easier to access and repair (or replace) components. While both STRAP and the SNL prototype are designed to simultaneously hold up to three radiometers with a maximum combined weight of 8 kg, STRAP is able to hold the three instruments level to better than ±0.02° for aircraft pitch and roll angles of up to approximately ±10° (as opposed to the ±5° range of the SNL prototype). This increased capacity to hold three radiometers level to a higher accuracy and for a wider range of pitch and roll angles significantly adds to the utility of STRAP and the quality and quantity of the measurements.

In this paper we will describe STRAP and illustrate the improvements attainable in aircraft based irradiance measurements by use of STRAP. In section 2 we elaborate on the difficulties in making irradiance measurements from a moving aircraft to explain our motivation in developing STRAP. We briefly discuss the conventional techniques that attempt to minimize and correct the errors in irradiance measurements from rigidly mounted aircraft radiometers and the problems with these techniques. In section 3 we describe the design, operation, and specifications of STRAP. In section 4 we illustrate the performance capabilities of STRAP during normal flight operations and during a series of flight maneuvers designed to test the accuracy, range, and robustness of the platform while mounted on the CIRPAS Twin Otter aircraft. A solar pyranometer was mounted on STRAP, and an identical solar pyranometer was rigidly mounted nearby directly to the aircraft. Comparisons between the measurements from the two solar pyranometers are used to illustrate the accuracy and capability of the new platform. Finally, in section 5 we summarize our results and discuss future work.

2. Problems and methods in aircraft irradiance measurements

The general definition of broadband irradiance, or flux density, F, is given as the amount of radiant energy incident on a unit area surface in unit time (usually expressed in watts per meter squared; Liou 1980):
i1520-0426-25-12-2161-e1
where I(θ, ϕ, λ) is the specific intensity, or radiance, at a given wavelength, λ, incident on the unit area from a given zenith, θ, and azimuth, ϕ, angle. The integration is carried out over the complete hemisphere, that is, 0°–90° in zenith and 0°–360° in azimuth (see Fig. 1). This general definition gives the irradiance with respect to the given unit area regardless of the orientation of that unit area. For climate- and weather-related studies, it is the amount of radiative energy deposited at a given altitude of the atmosphere that is of specific interest and this is given by the irradiance, FE, for a unit area that is horizontally level with respect to earth-centered coordinates (Fig. 1a). The goal of aircraft-mounted radiometers is to measure FE, but what they actually measure is FR, the irradiance with respect to the unit area that is horizontally level with respect to the radiometer (Fig. 1b). Since aircraft rarely fly level, a rigidly mounted radiometer will rarely be level with respect to earth-centered coordinates and the measured irradiance, FR, will usually differ from the desired irradiance, FE. The magnitude of this difference will vary depending on the degree of tilt of the radiometer and the angular distribution of the incoming radiation across the sky. The difference is minimal for a nearly level radiometer measuring a nearly isotropic radiation field, but the difference increases as the tilt of the radiometer becomes greater and the radiation field becomes less isotropic.
This is best illustrated by looking at the case where the problem is most apparent—the measurement of downwelling solar radiation. The downwelling sunlight at a given altitude can be thought of as consisting of two components: the radiation coming directly from the sun, Isolar,dir, and the diffuse sunlight, Isolar,diff, scattered from the complete hemisphere by the air (i.e., Rayleigh scattering) and particles in the atmosphere (clouds, dust, pollution, etc.). The aircraft-measured downwelling solar irradiance with respect to radiometer-centered coordinates, Fsolar,R, can therefore be expressed as an expansion of Eq. (1) into two terms, a direct and diffuse component:
i1520-0426-25-12-2161-e2
Here θR0 and ϕR0 are the zenith and azimuth angles, respectively, of the sun in radiometer-centered coordinates and the intensity has been integrated over all solar wavelengths (0.2–3.6 μm). The ratio of direct (term 1) to diffuse (term 2) radiation varies with sky conditions. For clear skies the direct radiation may be 90% or more of the total radiation, while for overcast skies, or conditions where the sun is obscured by clouds or a heavy aerosol loading, the direct component may be equal to the diffuse component or zero. However, in many situations the direct term is the significantly more dominant component of the measured irradiance.

For simplicity, therefore, let us assume that there is no diffuse component but only direct solar radiation [the first term only in Eq. (2)]. Such a situation could come close to occurring under very clear, clean skies at high altitude. It represents an extremely nonisotropic radiation field and will therefore most clearly illustrate the effect of the orientation of the radiometer on the measured signal. For this case the change in the measured solar irradiance scales directly with the change in the cosine of the solar zenith angle with respect to the radiometer. On an aircraft the cosine of the solar zenith angle with respect to the radiometer is not only a function of the position of the sun in the sky but also of the pitch, roll, and heading of the aircraft. Figure 2 illustrates how the solar zenith angle with respect to the radiometer can change solely because of a change in heading of the airplane. Figure 3 extends this example by showing the differences between the cosine of the solar zenith angle with respect to the radiometer and the cosine of the “true” solar zenith angle with respect to earth-centered coordinates, for a range of true solar zenith angles and aircraft pitch angles, and for headings toward and away from the sun. As shown by the first term in Eq. (2), such changes in solar zenith angle with respect to the radiometer will change the measured irradiance, even though the sun’s position in the sky and the direct component of the intensity, Isolar,dir, have not changed. With the addition of more isotropic diffuse radiation [the second term in Eq. (2)] this effect can be lessened, but since the direct solar radiation is usually the more dominant term it is typical to see significant changes in the measured solar irradiance generated by nothing more than heading changes in the aircraft.

This behavior is illustrated in Fig. 4, which shows measurements of the downwelling solar irradiance by a radiometer rigidly mounted to the CIRPAS Twin Otter as it flies at different headings but at a constant altitude. Since the sun’s position in the sky and atmospheric conditions have not changed significantly in the short time it takes for the aircraft to turn to a new heading (typically only a few minutes or less), the differences in measured solar irradiance shown in Fig. 4 are mainly due to the varying orientation of the sensor with respect to the sun at different headings.

To minimize this effect aircraft attempt to fly as level as possible during flight segments where the main goal is the acquisition of radiometer data. For example, to obtain an altitude profile of the irradiance a “stacked” flight pattern is typically flown where the aircraft sequentially flies straight and level legs at multiple altitudes over the same geographic line. The drawbacks to this type of flight profile are that the data are obtained at only a limited number of discrete altitude levels and it can take a long time to complete. Any given leg can be from a few minutes to tens of minutes long, and when the time for climbing, descending, and maneuvering are added in a complete altitude stack can take an hour or more to complete. During this time the atmospheric conditions can change significantly making it difficult to relate the data from the top and bottom of the stack.

In postprocessing of the data there are typically two steps performed to try and maximize the accuracy of aircraft irradiance measurements. The first step is to simply ignore all data that were obtained while the aircraft was outside of some arbitrary range of pitch and roll angles (e.g., Valero et al. 1997). Usually any measurements made for pitch and roll angles greater than ±2° to ±5° are not analyzed. Depending on the flight patterns this procedure can eliminate a large fraction of the radiometer data from any given flight.

The next step is the attempt to correct the remaining data back to a level platform. The correction methods in this step and the problems associated with them are too involved to be discussed in detail here, but as previously mentioned they typically involve two additional steps: 1) coordinate transformations that use the airplanes navigational information (time, latitude, longitude, altitude, pitch, roll, and heading) and solar ephemeris data to determine the zenith and azimuth angles of the sun with respect to the rigidly mounted radiometer in order to scale the direct component of the measured solar irradiance back to earth-centered coordinates (Hammer et al. 1991), and 2) iterative methods where the total measured irradiance (direct plus diffuse) is scaled by a range of offsets to the aircraft’s pitch and roll values until the optimum pitch and roll offset pair is found, which minimizes the differences between downwelling irradiances at heading changes (Bannehr and Glover 1991; Saunders et al. 1992; Bannehr and Schwiesow 1993; Valero et al. 1997; Boers et al. 1998). These correction methods can be difficult and time consuming because the problem is usually ill defined. For example, the correction methods assume that the pitch and roll of the aircraft match the pitch and roll of the sensor. However, if the mounted sensor is not level with respect to the latitudinal and longitudinal axes of the aircraft then there may be unknown offsets between the pitch and roll of the aircraft and the pitch and roll of the sensor that limit the effectiveness of these correction methods. In practice all attempts are made to align the instruments with the axes of the aircraft, or to measure the offsets, to minimize or correct for this effect. However, there are very often practical limitations on where and how the sensors can be mounted (e.g., because of limited space), and it can be difficult to accurately determine the latitudinal and longitudinal axes of the aircraft while on the ground since aircraft are rarely level when parked.

The correction methods are also ill defined because they require both knowledge of the direct to diffuse ratio of sky radiation, and the ability to associate changes in signal with changes in aircraft heading. In situations where sky conditions are very nonisotropic (e.g., scattered cloud fields above the aircraft) the direct to diffuse ratio will be constantly changing and unknown, and it may be impossible to isolate changes in signal due to heading changes from changes in signal due to varying sky conditions. In these cases the correction methods can simply not be applied.

While all of the various correction methods mentioned above can improve the data, it is not uncommon to find that even after their application the irradiance data are still not fully corrected. This is evidenced by continued shifts (from a few percent to 30%) in the corrected data with heading changes of the aircraft for all or some portion of a flight. In addition, even if the corrected irradiance does not vary with heading there is no way of knowing a priori whether the final, smooth value of the downwelling irradiance is systematically too high or too low.

3. Instrument description

The Stabilized Radiometer Platform (Fig. 5) was developed to address the difficulties discussed above in making and analyzing aircraft irradiance measurements. Its purpose is to keep a set of irradiance sensors level during flight. It was designed and built by Sonoma EO and incorporates off-the-shelf military and commercial technology with custom-made hardware designed by Sonoma EO.

STRAP consists of two major components (see Fig. 5), a controller and a gimbaled platform. The controller (on the left in Fig. 5) contains the electronics that monitor and control the actions of the gimbaled platform. It houses the control processor, GPS receiver, power filtering and conversion electronics, and electronic motor drives. The control processor is a PC-104 based computer system with motion control boards. An external monitor, keyboard, and mouse can be connected to the front of the controller allowing an operator to view the status and condition of the system. All user interface commands and data are transmitted by a serial RS-422 interface. The controller is a rugged chassis that fits in a standard 19-in. rack (dimensions: width 48.3 cm × height 17.8 cm × depth 38.1 cm; weight: 11 kg). It is powered from the 28 VDC avionics bus of the aircraft.

The gimbaled platform (on the right in Fig. 5) contains the hardware that actively stabilizes and levels the instruments. It is mounted on top of the aircraft by the 16 holes in the mounting plate. It consists of the payload plate assembly, an inertial measurement unit (IMU), a GPS antenna, a dynamic pressure shield (DPS; explained below), motors, actuators, potentiometers, and encoders (see Fig. 6). The radiometers are mounted on the top of the payload plate. The IMU is directly attached to the bottom of the payload plate (see Fig. 6a). The payload plate assembly, with attached radiometers and IMU, rotates in two axes relative to the aircraft (pitch and roll) and is the only portion of the entire assembly that is stabilized when the system is running. Figure 6b gives the dimensions of the gimbal. It weighs 30.4 kg.

This design differs from the stabilized system described by Wendisch et al. (2001), which has the IMU fixed to the aircraft and not to the sensor plate. The Wendisch et al. design reduces costs since only one IMU is needed to stabilize both an upward- and a downward-looking sensor, but it prevents a direct measurement of the attitude of the sensors and therefore limits the leveling accuracy. The STRAP was specifically designed for upward-looking sensors (i.e., the measurements most sensitive to attitude variations), and since the IMU is attached to the payload plate the attitude of the sensors can be measured directly to high accuracy.

In simple terms, STRAP works by sensing any change in both the attitude of the payload plate with respect to earth-centered coordinates, and the angular offsets between the payload plate and the aircraft, and then correcting for these changes through a series of actuators and motors. Fundamental to the operation of STRAP is the integration of a real-time inertial navigation system (INS). The INS senses the inertial motion of the payload plate to obtain its attitude. The system consists of a highly accurate IMU for rate and acceleration data, a GPS for positional information, and INS software for real-time analysis. As mentioned, the IMU is mounted directly to the payload plate assembly to sense its motion (see Fig. 6a). STRAP uses the Northrop Grumman tactical-grade LN-200 IMU with a stated attitude accuracy of better than 0.006°. It employs fiber-optic gyros (FOGs) to measure the payload plate angular rate and silicon accelerometers to measure linear acceleration. From this information the attitude (pitch, roll) of the payload plate assembly with respect to earth-centered coordinates is determined. A major source of error to positional stability originates from the inherent drift of the IMU. To compensate for this drift, a NovAtel GPS is used and integrated with the IMU to improve performance. The GPS antenna is mounted in close proximity to the payload plate assembly (see Fig. 5) to minimize the effects of aircraft structure bending with appropriate lever arm compensation for the distances and relationships between the IMU and GPS antenna. The platform control software incorporates Internav software by NavSys for analysis of the INS data. This is an embedded real-time application. The software is stored in nonvolatile memory on the controller and is designed to begin execution upon application of controller power.

To actively level the payload plate with respect to earth-centered coordinates the platform control software uses the attitude information obtained from the INS to command compensating motion of the payload plate assembly through a pair of axis motors (in pitch and roll). Sonoma EO designed the direct drive axis motors attached to the payload plate assembly that are used to accurately maintain its angular orientation. The software incorporates closed loop equations and Kalman filtering techniques to control the relative angle of the payload plate assembly to the airplane. Voltage outputs from the motion control cards are received by individual motor control driver modules, which then operate the axis motors. Encoders and potentiometers measure the positional relationships between payload plate and the aircraft that are fed back to the control processor for incorporation into the closed loop control algorithms.

Tests described below show that STRAP is able to actively keep the uplooking radiometers horizontally level to within ±0.02° for aircraft pitch and roll angles of up to approximately ±10°. The ±10° limit in pitch and roll is controlled by hardware stops in the gimbal. This limit ensures that the need for stabilization stays within the power range and capabilities of the axis motors. When the aircraft pitch or roll exceeds this limit the payload plate assembly gets buffered against the hardware stops and simply moves with the airplane, stabilizing back to level again when the pitch or roll comes back into the ±10° range.

During a flight the platform is continuously providing a 100-Hz stream of positional data to the user that includes: latitude, longitude, altitude, payload plate attitude (pitch, roll, and heading), aircraft to payload plate angles (pitch and roll), and INS uncertainty. These data are useful for detection of out-of-limit conditions such as aircraft pitch or roll in excess of 10°. The system update rate of 100 Hz compensates for most pitch and roll changes experienced in normal flight and in turbulence.

To minimize aerodynamic loading, the radiometers on STRAP are configured in line with the airstream, thereby reducing the projected width and streamlining the platform. In addition, the dynamic pressure shield (see Figs. 5 and 6b) is a small fairing that encircles the radiometers and minimizes the external destabilizing forces of the airstream on the platform. It enables a smaller platform with reduced power requirements. The DPS is a dynamic fairing that rotates only in the pitch axis. It tracks the pitch-axis motion of the payload plate assembly but is mechanically decoupled from it. A differential potentiometer measurement is made between the payload plate and the DPS to measure and correct for any tracking errors of the DPS to the payload plate. Since the DPS does not rotate in roll, the sensor ports in the DPS are slot shaped to accommodate the full range of motion (±10°) required by the roll axis of the payload plate assembly.

The payload plate was designed to hold three radiometers (Fig. 6a). The locations on either end of the payload plate were designed to hold modified Kipp & Zonen CM22 pyranometers (Kipp & Zonen 2004) to measure solar irradiance or CG4 pyrgeometers (Kipp & Zonen 2001) to measure IR irradiance. The ruggedized radiometers used here were modified and flight tested as part of the Sandia National Laboratories ARM–UAV program and were provided courtesy of SNL. The modifications to make these radiometers better suited for aircraft use included amplifying and digitizing the analog signals right at the sensor head, reducing the thermal mass of the body, and mounting the instruments in the airstream. Both the solar and IR radiometers have a hemispheric field of view, a 10-Hz data rate, and RS232 data format output. The solar radiometer has a wavelength bandpass of 0.2–3.6 μm, while the IR radiometer has a wavelength bandpass of 4.5–42 μm. The center location on the plate was specifically designed to hold a fiber-optic spectroradiometer (Pilewskie et al. 2003). An open tube runs vertically through the center of the entire gimbaled platform to allow passage of the fiber-optic cable from the spectroradiometer through the gimbal and out the bottom for entry into the aircraft. While the payload plate was specifically designed to hold these types of radiometers it is possible to mount other radiometers or instruments on it by using appropriately designed adapter plates, as long as the new radiometers are within the size (diameters <5.72 cm nominal) and weight (<8 kg cumulative weight of all sensors) restrictions of STRAP and the configuration is properly balanced.

4. Instrument performance

The Stabilized Radiometer Platform was mounted on the NPS CIRPAS Twin Otter aircraft for the DOE Aerosol Intensive Operations Period (IOP) field study in late April and May 2003 (Ferrare et al. 2006). Its performance was evaluated during normal flight operations and during a series of flight maneuvers designed to test the accuracy, range, and robustness of the platform. A modified Kipp & Zonen CM22 solar pyranometer (described in the previous section) was mounted on STRAP, and an identical modified Kipp & Zonen CM22 solar pyranometer was rigidly mounted nearby directly to the aircraft (Fig. 7). In this section, angular attitude data from the INS–GPS system on STRAP and comparisons between the measurements from the two solar pyranometers are used to illustrate the accuracy and capability of the new platform.

The INS–GPS system on STRAP measures both the angular attitude of the payload plate assembly with respect to earth-centered coordinates and the angular offsets between the payload plate and the aircraft. From these measured angular offsets the pitch and roll of the aircraft can be derived. In addition, since the IMU is directly mounted to the payload plate assembly (as described in section 3 above) it gives a direct measurement of the leveling accuracy of the platform.

Figure 8 shows a comparison between the pitch and roll of the aircraft and the pitch and roll of the payload plate assembly as measured by the INS–GPS system on STRAP. Data for the entire DOE Aerosol IOP flight on 12 May 2003 are presented. This flight included numerous turns, ascents, descents, spirals, and mild turbulence and illustrates the excellent stability that can be achieved by the platform during a typical scientific research flight. The horizontal black line through the center of the top two plots (Figs. 8a,b) does not represent the origin but is the measured pitch or roll of the payload plate assembly with respect to earth-centered coordinates. Figures 8c,d show a magnified view of the payload plate pitch and roll, respectively, as measured by the attached IMU. It can be seen that the payload plate remained level to better than ±0.02° for the majority of the flight.

Figure 8 also gives exact information on the pitch and roll limits of STRAP. As mentioned, the aircraft pitch and roll values displayed in Figs. 8a,b are derived from the INS-measured angular offsets between the payload plate and the aircraft. These values are only valid when the aircraft pitch and roll are within the limits of STRAP (approximately ±10°). Once the aircraft pitch or roll exceeds those limits the payload plate assembly becomes buffered against the hardware stops and moves with the airplane (stabilizing back to level again when the aircraft pitch or roll comes back into the ±10° range). This is why the aircraft pitch and roll values in Fig. 8 (red line) never exceed approximately ±10° even though the aircraft pitch or roll was much greater at times. In fact, the maximum and minimum values of the red lines in Fig. 8 actually represent the pitch and roll limits of the platform. Figure 8a shows that STRAP can compensate for airplane pitch-up angles of up to +10°. The airplane was almost always pitched up or level during this flight, so the maximum pitch down angle is not illustrated here. It was determined to be −10° on a subsequent flight. Figure 8b shows that STRAP can compensate for roll angles up to +10° in one direction (corresponding to the right, or starboard wing down) and −12° in the other direction (corresponding to the left, or port wing down). This asymmetry is a result of inherent clearance constraints within the platform. STRAP can therefore handle slightly steeper turns to the left.

The plots in Fig. 8 are typical and representative of the performance capability of STRAP. They also illustrate the vast increase in the amount of irradiance data that can be obtained from a flight. As mentioned in section 2, in the past, without the platform, all the zenith radiometer data where the pitch or roll of the airplane was greater than ±2 to ±5 degrees would have been ignored and not analyzed. As shown in Fig. 8, without the platform a good percentage of this flight would have yielded useless irradiance measurements. With the platform, good zenith irradiance data were obtained for practically the whole flight.

To fully evaluate STRAP, the CIRPAS Twin Otter performed a series of flight maneuvers designed to test the accuracy, range, and robustness of the platform. Figure 9 compares the downwelling solar irradiance measured by the radiometer mounted on STRAP (black line) to the solar irradiance measured by the radiometer rigidly mounted to the airplane (red line) for different portions of the 30 April 2003 DOE Aerosol IOP flight. We limit our discussion here to the solar irradiance measurements because, as explained in section 2, the solar radiation field is typically nonisotropic and dominated by the strong component of radiation coming directly from the sun. Aircraft measurements of the downwelling solar irradiance, therefore, tend to be most sensitive to the orientation of the sensor. Measurements of the IR irradiance are less sensitive because the IR radiation field is typically more diffuse and isotropic and less affected by the orientation of the instrument. The comparisons presented here between the rigidly mounted pyranometer and the level pyranometer were done under mostly clear skies and will therefore most rigorously test and highlight the capabilities of STRAP.

All of the maneuvers were done at a constant altitude of 3500 m. To maintain level flight during these flight segments the Twin Otter maintained a slightly pitched-up attitude of 5° unless otherwise noted. Sky conditions were mostly clear for the duration of these maneuvers; however, thin, scattered cirrus and some long-lived contrails were reported above the airplane, with no clouds reported below the airplane.

To test the performance of STRAP with changes in pitch and roll the Twin Otter first flew a heading directly toward the sun and the pitch of the airplane was rapidly varied between plus and minus 10° while keeping the roll level at approximately 0° (Fig. 9a). Next the Twin Otter flew a heading that kept the sun directly off the right (starboard) wing. The roll of the airplane was then rapidly varied between plus and minus 10° while keeping the pitch constant at approximately 5° (Fig. 9b). The purpose of these flight segments was to isolate and test the pitch and roll characteristics of the platform while keeping the airplane pitch and roll angles within the capabilities of the platform to remain level (approximately ±10°). As shown in Figs. 9a,b, during these maneuvers the solar irradiance measured by the rigidly mounted radiometer varied by approximately ±50 W m−2 and was systematically lower than the irradiance measured by the radiometer on the platform. In contrast, the solar irradiance measured by the radiometer on the platform did not vary with changes in pitch and roll. The reason for the systematically lower irradiances measured by the rigidly mounted radiometer is that the airplane was flying in a slightly nose-up attitude (except when performing the pitch maneuvers) and for headings into the sun, or with the sun off of the wing, this caused the solar zenith angle with respect to the rigidly mounted radiometer to be greater than the solar zenith angle with respect to the level radiometer on STRAP. As noted in section 2 and in Fig. 2, this will cause the irradiance measured by the rigidly mounted radiometer to be less than the irradiance measured by the level radiometer on STRAP. Figures 9a,b show that STRAP can keep radiometers level within the approximately ±10° desired range of airplane pitch and roll angles.

The next flight maneuver in the series tested the performance of STRAP with rapid changes in yaw. The airplane flew a heading directly into the sun again. The yaw of the Twin Otter was then rapidly varied between approximately ±10° while attempting to keep the pitch constant at approximately 5° and roll approximately level. This was difficult and there was some variation in the pitch and roll of the aircraft during this maneuver. Figure 9c shows that the solar irradiance measured by the radiometer on STRAP did not vary with changes in yaw (black line), while the solar irradiance measured by the rigidly mounted radiometer (red line) was systematically lower and showed some variation. As previously mentioned, the rigidly mounted radiometer measured a slightly lower irradiance because of the pitched-up attitude of the airplane. The variation in the irradiance from the rigidly mounted radiometer was due to variations in the pitch and roll of the aircraft during the yaw maneuvers. Figure 9c shows that STRAP can keep the radiometers level during changes in yaw of the aircraft.

The final flight maneuver tested the ability of STRAP to recover from airplane bank angles that exceeded its range of ±10° (Fig. 9d). The Twin Otter flew a heading away from the sun and performed a pair of back-to-back 360° (i.e., full circle) steep turns with a constant bank angle of approximately 30°. The first turn was to the left and the second turn was to the right. For this case the irradiance measured by the radiometer on STRAP varies during the turns because the platform rolls with the airplane once the bank exceeds approximately ±10°. However, Fig. 9d shows that the platform recovered immediately once the bank of the airplane came back within the ±10° range.

To highlight the utility of STRAP under cloudy-sky conditions data from a portion of the 12 May 2003 flight are shown in Fig. 10. During this portion of the flight scattered to mostly overcast cirrus clouds and persistent contrails were encountered. The flight segment shown includes a spiral ascent from approximately 396 to 2896 m (mission times: 4.50–4.85 h), a level leg with heading changes at 2896 m (mission times: 4.85–5.15 h), a slant descent to 1402 m (mission times: 5.15–5.25 h), and a level leg with heading changes at 1402 m (mission times: 5.25–5.5 h). Figure 10 shows that under these particularly complicated conditions the differences between the level and nonlevel solar radiometer measurements varied from approximately 0 to 100 W m−2. The magnitude of such differences depends on many factors such as the pitch, roll, and heading of the airplane, the solar zenith angle, and the aerosol and cloud conditions. All of these factors affect the direct–diffuse ratio making this type of sky condition especially troublesome for rigidly mounted aircraft radiometers because there is no way of knowing whether the changes in measured irradiance are due to changes in the orientation of the instrument or actual irradiance changes due to sky conditions. In fact, rigidly mounted aircraft solar radiometer measurements under such partly cloudy skies are virtually uncorrectable. The advantage of keeping the radiometers level, as done with STRAP, is that the measured irradiance under such sky conditions is the true downwelling irradiance. This ability of STRAP to accurately measure the solar irradiance from an aircraft under partly cloudy skies is a significant advancement in airborne radiometry.

The final illustration of the effectiveness of STRAP is shown in Fig. 11. During the DOE Aerosol IOP the Twin Otter performed numerous spiral ascents and descents. Figure 11 shows the downwelling solar irradiance measured at the beginning of the 12 May 2003 flight as the aircraft made a 152 m per minute spiral ascent from approximately 61 to 5334 m in altitude. Skies were clear during this portion of the flight. Throughout the spiral the pitch of the aircraft was approximately 7°–8° (i.e., nose-up) while the roll averaged approximately −5° (i.e., left, or port wing down), but varied from approximately −2° to −10°. Figure 11 illustrates the huge swings in the data (±100 W m−2) from the rigidly mounted radiometer compared to the steady data from the radiometer on the platform. In the past, none of the radiometer data in such a spiral would have been used because the pitch and roll angles of the aircraft would have been out of range. Now, using STRAP, accurate downwelling solar irradiances were obtained throughout the spiral profiles from near the surface to almost 5500 m. The advantages of this new ability to acquire irradiance data during a spiral, as opposed to multiple altitude straight and level legs, are that accurate measurements of the solar and IR irradiances, the flux divergence, and the heating rates can now be obtained for the complete vertical atmospheric column in a relatively short amount of time (approximately 20–40 min), and within a close radius to any surface measurements. For this particular case, the spiral had a diameter of approximately 8 km centered over the DOE Cloud and Radiation Testbed (CART) site. Relating the measurements from the bottom to the top of the vertical profile is therefore more appropriate and the data are more applicable to direct comparison with model-calculated data constrained by surface measurements.

As shown in this section, STRAP has the ability to perform extremely well and it was operational for the majority of the flights during the DOE field study. It did experience some hardware and software problems that caused it to fail during some of the subsequent flights; however, these problems were all repaired. In November of 2005 STRAP underwent further flight testing on the CIRPAS Twin Otter. For five flights, STRAP underwent a series of rigorous flight maneuvers similar to those described in this section that were designed to push the limits of its capabilities. It performed well for all of those flights. However, increasing the reliability of STRAP is still an issue being pursued. Through additional flight testing and participation in field studies it appears that STRAP currently tends to experience a component failure every 5–8 flights. Further testing and evaluation is planned and in progress in an ongoing effort to increase this reliability rate.

5. Summary and future work

A newly developed, actively stabilized, horizontally level platform that improves both the quality and quantity of aircraft measurements of the downwelling solar and infrared irradiance has been described. The Stabilized Radiometer Platform (STRAP) can hold up to three radiometers simultaneously and keep them level to better than ±0.02° for aircraft pitch and roll angles of up to approximately ±10°. STRAP was mounted on the CIRPAS Twin Otter aircraft and its performance evaluated during normal flight operations and during a series of flight maneuvers designed to test the accuracy, range, and robustness of the platform. Results were presented showing that STRAP kept the radiometers level within the specified pitch and roll range, that it was able to recover from flight maneuvers outside of this range, and that it greatly increased the quantity of useful radiometer data from any given flight. In particular, it was shown that STRAP is able to obtain accurate radiometer measurements of the downwelling solar irradiance during spiral ascents or descents, greatly expanding the utility of aircraft radiometer measurements.

STRAP addresses many of the problems associated with making solar and IR irradiance measurements using radiometers rigidly attached to an aircraft:

  1. By actively keeping the radiometers level during flight, STRAP eliminates the errors introduced into the measurements from a tilted instrument. The zenith angles with respect to the radiometer of the incoming radiation, whether direct or diffuse, are now identical to the zenith angles with respect to earth-centered coordinates. This means that the postprocessing techniques to correct the measured irradiance back to a level platform (coordinate transformations, iterative schemes) are no longer needed. While STRAP does not correct for errors in the sensors themselves (such as a nonideal cosine response or a nonlevel receiving plane) it greatly simplifies the data-reduction process. With the application of the proper calibration constants an accurate estimate of the irradiance could even be obtained in real time in flight.

  2. STRAP has its own self-contained INS–GPS navigational system attached directly to the platform holding the radiometers. This eliminates any error introduced into the irradiance measurements due to an offset between the pitch and roll of the instruments and the pitch and roll of the aircraft. All that is required is that the radiometers be mounted level on STRAP, a much simpler procedure than trying to mount them level to the latitudinal and longitudinal axes of the airplane.

  3. Since STRAP keeps the radiometers accurately level for pitch and roll angles of up to approximately ±10° the restrictions on flight patterns are greatly reduced and valid data can be obtained even in spiral ascents (Fig. 11) and descents. This adds greatly to the quantity of data from any given flight since in many cases almost all of the maneuvering can be accomplished by keeping the pitch and roll of the airplane within the ±10° range. Also, fast profiling of the radiative balance of the atmospheric column is now possible.

As shown in this paper STRAP greatly increases the quality and quantity of aircraft radiometer irradiance measurements and greatly simplifies the data-reduction process. Future work on STRAP will entail efforts to increase the robustness and reliability of the platform. Beyond this, STRAP will be used to address scientific questions requiring accurate airborne radiometric measurements that have been problematic in the past. For example, for the first time, STRAP provides an opportunity to truly evaluate current methods used to correct airborne irradiance measurements from rigidly mounted radiometers. Work is progressing on a more in depth analysis of data similar to that presented in this paper comparing irradiance measurements from level radiometers mounted on STRAP with irradiance measurements from radiometers rigidly attached to the aircraft. In addition, STRAP will be used in field studies investigating the direct and indirect effects of aerosols on the radiative balance of the earth’s atmosphere.

Acknowledgments

The authors gratefully acknowledge the efforts and support of the pilots, technicians, and staff at the Center for Interdisciplinary Remotely Piloted Aircraft Studies. In particular we wish to thank Dr. Haflidi H. Jonsson, chief scientist; Mike Hubbell, pilot; Roy Woods, pilot/engineer; Reggie Burch, avionics mechanic; and Nava Roy, systems programmer. The development of STRAP was funded by an Office of Naval Research (ONR) Small Business Innovation Research (SBIR) grant. The support of ONR Program Element PE061153N is gratefully acknowledged.

REFERENCES

  • Ackerman, S. A., and Cox S. K. , 1981: Aircraft observations of the shortwave fractional absorptance of non-homogeneous clouds. J. Appl. Meteor., 20 , 15101515.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Albrecht, B. A., Poellet M. , and Cox S. K. , 1974: Pyrgeometer measurements from aircraft. Rev. Sci. Instrum., 45 , 3338.

  • Austen, I., 2002: Jitter-free, but not worry-free. New York Times, 20 June, Thursday ed., G1, G4.

  • Bannehr, L., and Glover V. , 1991: Preprocessing of airborne pyranometer data. NCAR Tech. Note NCAR/TN-364+STR, 35 pp.

  • Bannehr, L., and Schwiesow R. , 1993: A technique to account for the misalignment of pyranometers installed on aircraft. J. Atmos. Oceanic Technol., 10 , 774777.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Boers, R., Mitchell R. M. , and Krummel P. B. , 1998: Correction of aircraft pyranometer measurements for diffuse radiance and alignment errors. J. Geophys. Res., 103 , 1675316758.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Cox, S. K., and Griffith K. T. , 1979: Estimates of radiative divergence during phase III of the GARP Atlantic Tropical Experiment: Part I. Methodology. J. Atmos. Sci., 36 , 576585.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Ferrare, R., Feingold G. , Ghan S. , Ogren J. , Schmid B. , Schwartz S. E. , and Sheridan P. , 2006: Preface to special section: Atmospheric Radiation Measurement Program May 2003 intensive operations period examining aerosol properties and radiative influences. J. Geophys. Res., 111 .D05S01, doi:10.1029/2005JD006908.

    • Search Google Scholar
    • Export Citation
  • Fritz, S., 1948: The albedo of the ground and atmosphere. Bull. Amer. Meteor. Soc., 29 , 303312.

  • Hammer, P. D., Valero F. P. J. , and Kinne S. , 1991: The 27–28 October 1986 FIRE cirrus case study: Retrieval of cloud particle sizes and optical depths from comparative analyses of aircraft and satellite-based infrared measurements. Mon. Wea. Rev., 119 , 16731692.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Hayasaka, T., Kikuchi N. , and Tanaka M. , 1995: Absorption of solar radiation by stratocumulus clouds: Aircraft measurements and theoretical calculations. J. Appl. Meteor., 34 , 10471055.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Hignett, P., Taylor J. P. , Francis P. N. , and Glew M. D. , 1999: Comparison of observed and modeled direct aerosol forcing during TARFOX. J. Geophys. Res., 104 , 22792287.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Jäkel, E., Wendisch M. , Kniffka A. , and Trautmann T. , 2005: Airborne system for fast measurements of upwelling and downwelling spectral actinic flux densities. Appl. Opt., 44 , 434444.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Kipp & Zonen, 2001: CG 4 pyrgeometer instruction manual. Kipp & Zonen, 64 pp. [Available online at http://www.kippzonen.com/?downloadcategory/38192/Discontinued+Solar+Instruments.aspx.].

  • Kipp & Zonen, 2004: CM 22 precision pyranometer instruction manual. Kipp & Zonen, 65 pp. [Available online at http://www.kippzonen.com/?downloadcategory/38192/Discontinued+Solar+Instruments.aspx.].

  • Kuhn, P. M., and Suomi V. E. , 1958: Airborne observations of albedo with a beam reflector. J. Meteor., 15 , 172174.

  • Liou, K-N., 1980: An Introduction to Atmospheric Radiation. International Geophysics Series, Vol. 26, Academic Press, 392 pp.

  • Pilewskie, P., and Coauthors, 2003: Solar spectral radiative forcing during the Southern African Regional Science Initiative. J. Geophys. Res., 108 .8486, doi:10.1029/2002JD002411.

    • Search Google Scholar
    • Export Citation
  • Rawlins, F., 1989: Aircraft measurements of the solar absorption by broken cloud fields: A case study. Quart. J. Roy. Meteor. Soc., 115 , 365382.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Roach, W. T., 1961: Some aircraft observations of fluxes of solar radiation in the atmosphere. Quart. J. Roy. Meteor. Soc., 87 , 346363.

  • Saunders, R. W., Brogniez G. , Buriez J. C. , Meerkotter R. , and Wendling P. , 1992: A comparison of measured and modeled broadband fluxes from aircraft data during the ICE ’89 field experiment. J. Atmos. Oceanic Technol., 9 , 391406.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Valero, F. P. J., and Coauthors, 1997: Atmospheric Radiation Measurements Enhanced Shortwave Experiment (ARESE): Experimental and data details. J. Geophys. Res., 102 , 2992929937.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Valero, F. P. J., and Coauthors, 2003: Absorption of solar radiation by the clear and cloudy atmosphere during the Atmospheric Radiation Measurement Enhanced Shortwave Experiments (ARESE) I and II: Observations and models. J. Geophys. Res., 108 .4016, doi:10.1029/2001JD001384.

    • Search Google Scholar
    • Export Citation
  • Wendisch, M., and Keil A. , 1999: Discrepancies between measured and modeled solar and UV radiation within polluted boundary layer clouds. J. Geophys. Res., 104 , 2737327385.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Wendisch, M., and Mayer B. , 2003: Vertical distribution of spectral solar irradiance in the cloudless sky: A case study. J. Geophys. Res., 30 .1183, doi:10.1029/2002GL016529.

    • Search Google Scholar
    • Export Citation
  • Wendisch, M., Müller D. , Schell D. , and Heintzenberg J. , 2001: An airborne spectral albedometer with active horizontal stabilization. J. Atmos. Oceanic Technol., 18 , 18561866.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Wendisch, M., and Coauthors, 2004: Airborne measurements of areal spectral surface albedo over different sea and land surfaces. J. Geophys. Res., 109 .D08203, doi:10.1029/2003JD004392.

    • Search Google Scholar
    • Export Citation

Fig. 1.
Fig. 1.

Illustration of the terms used in the definition of irradiance as given by Eq. (1) in the text. The irradiance is always defined with respect to a unit surface area and is not dependent on the orientation of that unit area (on the left). (a) For atmospheric studies, it is the irradiance measured by a horizontally level radiometer F that is desired, but (b) since aircraft are rarely level in flight, it is the irradiance measured from a tilted radiometer FR that is typically obtained.

Citation: Journal of Atmospheric and Oceanic Technology 25, 12; 10.1175/2008JTECHA1085.1

Fig. 2.
Fig. 2.

Illustration of the effect of solar zenith angle, aircraft pitch, and aircraft heading on the zenith angle of the sun with respect to the aircraft normal. In both pictures the solar zenith angle with respect to earth-centered coordinates is approximately 45°, and the aircraft is pitched up at 20° (an exaggerated pitch angle for illustrative purposes). However, the solar zenith angle with respect to the aircraft normal is greater when (left) heading toward the sun (θ1) than when (right) heading away from the sun (θ2), even though the true solar zenith angle has not changed. Since irradiance sensors have a cosine angular response, radiometers rigidly mounted on a pitched-up aircraft will measure a direct solar irradiance that is less when heading into the sun than when heading away from the sun.

Citation: Journal of Atmospheric and Oceanic Technology 25, 12; 10.1175/2008JTECHA1085.1

Fig. 3.
Fig. 3.

Percent difference between the cosine of the solar zenith angle with respect to the aircraft normal (θR) and the cosine of the true solar zenith angle with respect to earth-centered coordinates (θE) as a function of true solar zenith angle and aircraft pitch and heading. The measured solar irradiance for a heading toward the sun would be less than for a heading away from the sun.

Citation: Journal of Atmospheric and Oceanic Technology 25, 12; 10.1175/2008JTECHA1085.1

Fig. 4.
Fig. 4.

Typical example of the systematic change in the measured downwelling solar irradiance with changes in heading when using radiometers rigidly mounted to the aircraft. These data are from a flight of the CIRPAS Twin Otter on 12 May 2003 as part of the DOE Aerosol IOP field study. The Twin Otter remained at a constant altitude for this whole flight segment. Mission time is simply defined as the time since the start of the onboard data acquisition system at the beginning of the flight.

Citation: Journal of Atmospheric and Oceanic Technology 25, 12; 10.1175/2008JTECHA1085.1

Fig. 5.
Fig. 5.

Picture of the STRAP showing its two major components: the controller on the left that is mounted inside the aircraft on a standard 19-in. rack and the gimbal on the right that is mounted on top of the aircraft. (Photo courtesy of Roy Woods)

Citation: Journal of Atmospheric and Oceanic Technology 25, 12; 10.1175/2008JTECHA1085.1

Fig. 6.
Fig. 6.

Computer-aided design drawing of the interior of the gimbal to show key components: (a) with both the outer fairing and DPS removed; and (b) with only the outer fairing removed.

Citation: Journal of Atmospheric and Oceanic Technology 25, 12; 10.1175/2008JTECHA1085.1

Fig. 7.
Fig. 7.

Close-up of the radiometers mounted on STRAP on the CIRPAS Twin Otter for the DOE Aerosol IOP in May 2003. The radiometer on the far left is a modified Kipp & Zonen CM22 pyranometer and the radiometer on the far right is a modified Kipp & Zonen CG4 pyrgeometer. The instrument in the middle is the Solar Spectral Flux Radiometer (SSFR) from NASA Ames Research Center (Pilewskie et al. 2003). In the background, just behind the SSFR, is a fourth radiometer mounted on a black tower. This is also a modified Kipp & Zonen CM22 pyranometer rigidly mounted directly to the aircraft for comparison with the measurements from the CM22 on STRAP.

Citation: Journal of Atmospheric and Oceanic Technology 25, 12; 10.1175/2008JTECHA1085.1

Fig. 8.
Fig. 8.

Illustration of the stability and accuracy of the stabilized radiometer platform. (top) Comparison of the (a) pitch and (b) roll of the aircraft (red lines) with the pitch and roll of STRAP (black lines) while mounted on the CIRPAS Twin Otter for the entire DOE Aerosol IOP flight on 12 May 2003. (bottom) Magnified views of the platform (c) pitch and (d) roll showing it remained level to better than ±0.02° for most of the flight.

Citation: Journal of Atmospheric and Oceanic Technology 25, 12; 10.1175/2008JTECHA1085.1

Fig. 9.
Fig. 9.

Illustration of the ability of STRAP to remain level during flight maneuvers that sequentially varied the aircraft (a) pitch, (b) roll, and (c) yaw; and (d) to recover from bank angles that exceeded its approximately ±10° range of stabilization. The black lines are the irradiance measured by the radiometer on STRAP, while the red lines are the irradiance measured by the radiometer rigidly mounted to the aircraft. The airplane flew a constant altitude pattern for all maneuvers.

Citation: Journal of Atmospheric and Oceanic Technology 25, 12; 10.1175/2008JTECHA1085.1

Fig. 10.
Fig. 10.

Comparison of solar irradiance measurements under cirrus clouds for the 12 May 2003 flight of the CIRPAS Twin Otter during the DOE Aerosol IOP. The variations shown in the downwelling solar irradiance measured by the radiometer rigidly mounted to the aircraft (red line) are due to the variable cloud conditions (scattered to overcast cirrus, contrails) and to the changing orientation of the instrument with pitch, roll, and heading changes of the aircraft. Under such variable cloud conditions it would be difficult, if not impossible, to adequately correct the data from this radiometer in order to isolate the signal that is due to the clouds from that due to the varying orientation of the instrument. Since the radiometer on STRAP remained level, the variation in solar irradiance that it measured (black line) is only due to the variation in the cloud conditions.

Citation: Journal of Atmospheric and Oceanic Technology 25, 12; 10.1175/2008JTECHA1085.1

Fig. 11.
Fig. 11.

Comparison of the downwelling solar irradiance measured during a spiral ascent of the CIRPAS Twin Otter during the DOE Aerosol IOP flight on 12 May 2003. The black line is the irradiance measured by the radiometer on STRAP, while the red line is the irradiance measured by the radiometer rigidly mounted to the aircraft.

Citation: Journal of Atmospheric and Oceanic Technology 25, 12; 10.1175/2008JTECHA1085.1

Save
  • Ackerman, S. A., and Cox S. K. , 1981: Aircraft observations of the shortwave fractional absorptance of non-homogeneous clouds. J. Appl. Meteor., 20 , 15101515.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Albrecht, B. A., Poellet M. , and Cox S. K. , 1974: Pyrgeometer measurements from aircraft. Rev. Sci. Instrum., 45 , 3338.

  • Austen, I., 2002: Jitter-free, but not worry-free. New York Times, 20 June, Thursday ed., G1, G4.

  • Bannehr, L., and Glover V. , 1991: Preprocessing of airborne pyranometer data. NCAR Tech. Note NCAR/TN-364+STR, 35 pp.

  • Bannehr, L., and Schwiesow R. , 1993: A technique to account for the misalignment of pyranometers installed on aircraft. J. Atmos. Oceanic Technol., 10 , 774777.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Boers, R., Mitchell R. M. , and Krummel P. B. , 1998: Correction of aircraft pyranometer measurements for diffuse radiance and alignment errors. J. Geophys. Res., 103 , 1675316758.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Cox, S. K., and Griffith K. T. , 1979: Estimates of radiative divergence during phase III of the GARP Atlantic Tropical Experiment: Part I. Methodology. J. Atmos. Sci., 36 , 576585.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Ferrare, R., Feingold G. , Ghan S. , Ogren J. , Schmid B. , Schwartz S. E. , and Sheridan P. , 2006: Preface to special section: Atmospheric Radiation Measurement Program May 2003 intensive operations period examining aerosol properties and radiative influences. J. Geophys. Res., 111 .D05S01, doi:10.1029/2005JD006908.

    • Search Google Scholar
    • Export Citation
  • Fritz, S., 1948: The albedo of the ground and atmosphere. Bull. Amer. Meteor. Soc., 29 , 303312.

  • Hammer, P. D., Valero F. P. J. , and Kinne S. , 1991: The 27–28 October 1986 FIRE cirrus case study: Retrieval of cloud particle sizes and optical depths from comparative analyses of aircraft and satellite-based infrared measurements. Mon. Wea. Rev., 119 , 16731692.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Hayasaka, T., Kikuchi N. , and Tanaka M. , 1995: Absorption of solar radiation by stratocumulus clouds: Aircraft measurements and theoretical calculations. J. Appl. Meteor., 34 , 10471055.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Hignett, P., Taylor J. P. , Francis P. N. , and Glew M. D. , 1999: Comparison of observed and modeled direct aerosol forcing during TARFOX. J. Geophys. Res., 104 , 22792287.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Jäkel, E., Wendisch M. , Kniffka A. , and Trautmann T. , 2005: Airborne system for fast measurements of upwelling and downwelling spectral actinic flux densities. Appl. Opt., 44 , 434444.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Kipp & Zonen, 2001: CG 4 pyrgeometer instruction manual. Kipp & Zonen, 64 pp. [Available online at http://www.kippzonen.com/?downloadcategory/38192/Discontinued+Solar+Instruments.aspx.].

  • Kipp & Zonen, 2004: CM 22 precision pyranometer instruction manual. Kipp & Zonen, 65 pp. [Available online at http://www.kippzonen.com/?downloadcategory/38192/Discontinued+Solar+Instruments.aspx.].

  • Kuhn, P. M., and Suomi V. E. , 1958: Airborne observations of albedo with a beam reflector. J. Meteor., 15 , 172174.

  • Liou, K-N., 1980: An Introduction to Atmospheric Radiation. International Geophysics Series, Vol. 26, Academic Press, 392 pp.

  • Pilewskie, P., and Coauthors, 2003: Solar spectral radiative forcing during the Southern African Regional Science Initiative. J. Geophys. Res., 108 .8486, doi:10.1029/2002JD002411.

    • Search Google Scholar
    • Export Citation
  • Rawlins, F., 1989: Aircraft measurements of the solar absorption by broken cloud fields: A case study. Quart. J. Roy. Meteor. Soc., 115 , 365382.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Roach, W. T., 1961: Some aircraft observations of fluxes of solar radiation in the atmosphere. Quart. J. Roy. Meteor. Soc., 87 , 346363.

  • Saunders, R. W., Brogniez G. , Buriez J. C. , Meerkotter R. , and Wendling P. , 1992: A comparison of measured and modeled broadband fluxes from aircraft data during the ICE ’89 field experiment. J. Atmos. Oceanic Technol., 9 , 391406.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Valero, F. P. J., and Coauthors, 1997: Atmospheric Radiation Measurements Enhanced Shortwave Experiment (ARESE): Experimental and data details. J. Geophys. Res., 102 , 2992929937.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Valero, F. P. J., and Coauthors, 2003: Absorption of solar radiation by the clear and cloudy atmosphere during the Atmospheric Radiation Measurement Enhanced Shortwave Experiments (ARESE) I and II: Observations and models. J. Geophys. Res., 108 .4016, doi:10.1029/2001JD001384.

    • Search Google Scholar
    • Export Citation
  • Wendisch, M., and Keil A. , 1999: Discrepancies between measured and modeled solar and UV radiation within polluted boundary layer clouds. J. Geophys. Res., 104 , 2737327385.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Wendisch, M., and Mayer B. , 2003: Vertical distribution of spectral solar irradiance in the cloudless sky: A case study. J. Geophys. Res., 30 .1183, doi:10.1029/2002GL016529.

    • Search Google Scholar
    • Export Citation
  • Wendisch, M., Müller D. , Schell D. , and Heintzenberg J. , 2001: An airborne spectral albedometer with active horizontal stabilization. J. Atmos. Oceanic Technol., 18 , 18561866.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Wendisch, M., and Coauthors, 2004: Airborne measurements of areal spectral surface albedo over different sea and land surfaces. J. Geophys. Res., 109 .D08203, doi:10.1029/2003JD004392.

    • Search Google Scholar
    • Export Citation
  • Fig. 1.

    Illustration of the terms used in the definition of irradiance as given by Eq. (1) in the text. The irradiance is always defined with respect to a unit surface area and is not dependent on the orientation of that unit area (on the left). (a) For atmospheric studies, it is the irradiance measured by a horizontally level radiometer F that is desired, but (b) since aircraft are rarely level in flight, it is the irradiance measured from a tilted radiometer FR that is typically obtained.

  • Fig. 2.

    Illustration of the effect of solar zenith angle, aircraft pitch, and aircraft heading on the zenith angle of the sun with respect to the aircraft normal. In both pictures the solar zenith angle with respect to earth-centered coordinates is approximately 45°, and the aircraft is pitched up at 20° (an exaggerated pitch angle for illustrative purposes). However, the solar zenith angle with respect to the aircraft normal is greater when (left) heading toward the sun (θ1) than when (right) heading away from the sun (θ2), even though the true solar zenith angle has not changed. Since irradiance sensors have a cosine angular response, radiometers rigidly mounted on a pitched-up aircraft will measure a direct solar irradiance that is less when heading into the sun than when heading away from the sun.

  • Fig. 3.

    Percent difference between the cosine of the solar zenith angle with respect to the aircraft normal (θR) and the cosine of the true solar zenith angle with respect to earth-centered coordinates (θE) as a function of true solar zenith angle and aircraft pitch and heading. The measured solar irradiance for a heading toward the sun would be less than for a heading away from the sun.

  • Fig. 4.

    Typical example of the systematic change in the measured downwelling solar irradiance with changes in heading when using radiometers rigidly mounted to the aircraft. These data are from a flight of the CIRPAS Twin Otter on 12 May 2003 as part of the DOE Aerosol IOP field study. The Twin Otter remained at a constant altitude for this whole flight segment. Mission time is simply defined as the time since the start of the onboard data acquisition system at the beginning of the flight.

  • Fig. 5.

    Picture of the STRAP showing its two major components: the controller on the left that is mounted inside the aircraft on a standard 19-in. rack and the gimbal on the right that is mounted on top of the aircraft. (Photo courtesy of Roy Woods)

  • Fig. 6.

    Computer-aided design drawing of the interior of the gimbal to show key components: (a) with both the outer fairing and DPS removed; and (b) with only the outer fairing removed.

  • Fig. 7.

    Close-up of the radiometers mounted on STRAP on the CIRPAS Twin Otter for the DOE Aerosol IOP in May 2003. The radiometer on the far left is a modified Kipp & Zonen CM22 pyranometer and the radiometer on the far right is a modified Kipp & Zonen CG4 pyrgeometer. The instrument in the middle is the Solar Spectral Flux Radiometer (SSFR) from NASA Ames Research Center (Pilewskie et al. 2003). In the background, just behind the SSFR, is a fourth radiometer mounted on a black tower. This is also a modified Kipp & Zonen CM22 pyranometer rigidly mounted directly to the aircraft for comparison with the measurements from the CM22 on STRAP.

  • Fig. 8.

    Illustration of the stability and accuracy of the stabilized radiometer platform. (top) Comparison of the (a) pitch and (b) roll of the aircraft (red lines) with the pitch and roll of STRAP (black lines) while mounted on the CIRPAS Twin Otter for the entire DOE Aerosol IOP flight on 12 May 2003. (bottom) Magnified views of the platform (c) pitch and (d) roll showing it remained level to better than ±0.02° for most of the flight.

  • Fig. 9.

    Illustration of the ability of STRAP to remain level during flight maneuvers that sequentially varied the aircraft (a) pitch, (b) roll, and (c) yaw; and (d) to recover from bank angles that exceeded its approximately ±10° range of stabilization. The black lines are the irradiance measured by the radiometer on STRAP, while the red lines are the irradiance measured by the radiometer rigidly mounted to the aircraft. The airplane flew a constant altitude pattern for all maneuvers.

  • Fig. 10.

    Comparison of solar irradiance measurements under cirrus clouds for the 12 May 2003 flight of the CIRPAS Twin Otter during the DOE Aerosol IOP. The variations shown in the downwelling solar irradiance measured by the radiometer rigidly mounted to the aircraft (red line) are due to the variable cloud conditions (scattered to overcast cirrus, contrails) and to the changing orientation of the instrument with pitch, roll, and heading changes of the aircraft. Under such variable cloud conditions it would be difficult, if not impossible, to adequately correct the data from this radiometer in order to isolate the signal that is due to the clouds from that due to the varying orientation of the instrument. Since the radiometer on STRAP remained level, the variation in solar irradiance that it measured (black line) is only due to the variation in the cloud conditions.

  • Fig. 11.

    Comparison of the downwelling solar irradiance measured during a spiral ascent of the CIRPAS Twin Otter during the DOE Aerosol IOP flight on 12 May 2003. The black line is the irradiance measured by the radiometer on STRAP, while the red line is the irradiance measured by the radiometer rigidly mounted to the aircraft.

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 358 91 9
PDF Downloads 219 92 4