AIRSAR Operations Overview

 
 
 

Functionally, a typical AIRSAR operations can be divided into four main parts: (1) mission planning, (2) instrument maintenance and upgrades, (3) mission and flight operations, and (4) data processing and distribution. In reality, mission planning for the next year (and to some extent flight operations for the next year) takes place at the same time as data processing and distribution for the previous flight season. In this chapter a general discussion of AIRSAR operations is given.
 

2-1 Flight Request/Approval

AIRSAR flights are performed in response to flight requests submitted by the Principal Investigators (PIs) of NASA-sponsored scientific investigations and by sponsors from other organizations who contract to acquire AIRSAR data on a reimbursable basis. Appendix X contains a blank flight request, copies of which are available from the Aircraft Programs office at Ames Research Center in Mountain View, CA. A completed sample flight request is shown in Figure 2.X. Upon completion, flight requests should be sent to Ames Research Center and the relevant Program Manager at NASA Headquarters. Each year, usually in June, the cognizant NASA Headquarters Program Office reviews flight requests. Once NASA approves flight requests, Ames Research Center creates a mission schedule. Flight requests are typically due at Ames Research Center at the end of June each year, but a longer lead time may be necessary when a mission involves basing the DC-8 away from Moffett Field for extended periods.
 

2-2 Mission Planning

Since the DC-8 is a general-purpose flying laboratory, a number of different instruments are scheduled to use the aircraft every year. Assigning the times at which different instruments can be accomodated on the DC-8 is the responsibility of NASA and Ames. Once a time period for the AIRSAR system has been allocated and flight requests have been approved, mission planning for the upcoming season begins.

The primary purpose of the mission planning phase is to ensure that both the utilization of the AIRSAR hardware and the DC-8 is maximized, based on the flight requests. This process involves extensive discussions between the AIRSAR program and the Mission Managers at Ames, although the the primary responsibility for defining a mission lies with Ames.

Once a mission has been planned and the flight hours have been approved by the various program offices at NASA, specific flight lines for the various experiments are defined by the AIRSAR Experiment Scientist. These flight lines are created using information from the flight requests, as well as information obtained through direct contact with the PIs. Pre-mission operations are considered complete when a data acquisition program, including AIRSAR flight lines, has been defined. This plan is then forwarded to the navigators at Ames.

Home of future sample mission plan from flight covering S. Ste. Marie (flight lines I distribute at the start of each flight).

Ideally a data acquisition plan is created approximately one month before the first flight of an AIRSAR campaign. To achieve this goal, mission planning at JPL must begin at least two to three months before the season's first flight. Therefore, it is crucial that investigators provide their inputs in a timely manner to allow the AIRSAR personnel sufficient time to plan the flight operations of a campaign.
 

2-3 AIRSAR Installation

When not being used to acquire data, the AIRSAR system is maintained in a laboratory at JPL. Approximately two weeks before the start of a campaign, the AIRSAR system is shipped to Ames at Moffett Field, California, for installation on a DC-8 aircraft. The aircraft is maintained and operated by personnel at Ames.

Description of DC-8

NASA's Douglas DC-8-72 Aircraft is a four-jet engine aircraft, extensively modified as a flying laboratory. It has a range in excess of 5,000 nautical miles and a ceiling of 41,000 feet (12,500 meters). (AIRSAR flights are typically flown at an altitude of 26,000 feet and last eight or fewer hours.) To support various research missions there are facilities for power, intercom and air-to-ground communications, aerial photography and video, instrumentation racks, and a digital data stream of ephemeris data. Custom mounted hardware options are available both inside the cabin and on the skin of the aircraft. More information regarding the DC-8 can be found in the DC-8 Airborne Laboratory Experimenters Handbook, available from Ames research Center. Also, a staff of design engineers and a full sheet-metal and machine shop exists to support modifications and installations on the DC-8.
 

Installation on the DC-8

Approximately three weeks before the start of a campaign, a pre-ship review is held at JPL to insure that the AIRSAR system is ready to acquire data during the upcoming flights. After the pre-ship review, the AIRSAR instrument is transported to Ames Research Center. Once at Ames, the racks containing the AIRSAR equipment are weighed and inspected. A crew from the sheet metal shop then mounts the racks inside the DC-8, as well as installing the external antennas. Finally, the cables connecting the AIRSAR system are put in place by a member of the AIRSAR crew.

At the end of the flight season the AIRSAR flight crew, with the help of Ames personnel, removes the AIRSAR hardware from the aircraft. The hardware is then packed into the AIRSAR trailer and driven back to JPL where it is once again unpacked. The hardware is maintained and repaired in a laboratory at JPL.
 

2-4 Flight Operations

One of the most important phases of an AIRSAR campaign is the day to day flight operations when the AIRSAR data are acquired. For any given day a strawman flight plan is defined by the AIRSAR Experiment Scientist and passed on to the Ames navigators about two days before the actual flight is scheduled to take place. This lead time allows the navigators to generate an official flight plan and to file it with the relevant authorities for approval. In some cases, such as on deployments away from Moffett Field (and especially on deployments in foreign countries) flight plans must be filed well in advance of the actual flight date and last minute changes are not possible.

A flight plan for a typical day consists of a number of runs, lasting several hours. Each run is described by a flight line consisting of a start way point, a stop way point, and an altitude. The way points are defined in terms of the DC-8 ground track, so data acquired over the same site at different incidence angles will have different way points for each run. Also, the altitude requested is based on the elevation of the site and the requested incidence angle. In almost all cases the flight plan also includes way points that describe a route the DC-8 will fly from one site to another. Clearly, the flight plan is a very important document as it is used by the DC-8 pilots, the Ames mission managers, and the AIRSAR flight crew as their work plan for that day. Figure 2.X (map of S. Ste Marie with flight lines, not image swaths, with waypoints labeled ) shows the lines flown on a typical day with the way points labeled.

As mentioned above, a flight plan consists of a number of different runs. An AIRSAR run usually starts two and one half minutes before a point site and stops two and one half minutes after the point site, though longer runs are often flown. Data is collected in this manner to ensure that the plane is flying straight and level by the time it passes over the site. Once the recorders start up there are a few seconds of calibration data, including pre-chirps and receiver noise only, followed by the actual AIRSAR data. From 1988-1992 data were recorded on one of three High Density Digital Tape Recorders. Data were recorded at 10 MBytes per second on the HDDT and one HDDT can hold 2Gbytes of data. Since 1993, data are recorded on Sony tape recorders at the rate of 32 MBytes per second. One cassette tape on the Sony recorder can hold 96 GBytes of data. The high density digital tapes and Sony cassetes (HDDCs) are extremely valuable since they are the only copies of the raw data recorded.

In addition to normal AIRSAR data acquistion, baseline test data are also recorded on a regular basis, usually once each day on which data are recorded. These test data are later analyzed to ensure that the AIRSAR system is performing nominally.

The AIRSAR flight operations team consists of four positions: the radar operator, the support engineer, the correlator operator, and the experiment scientist. The radar operator sets up the radar parameters prior to a run, starts and stops the tape, and insures the data quality and general health of the system. The support engineer turns the transmitters on and off, changes the antenna routing depending on the data acquisition mode, and is responsible for fixing any hardware problems. The correlator operator monitors the data as it is being recorded and generates the survey prints after a data tape is completed. The experiment scientist is responsible for insuring that the flight plan generated prior to a flight is executed.

Flight operations are complete when all the tapes containing raw data have been shipped to JPL.
 

2-5 Data Processing

The purpose of this section is to give the user an understanding of the steps involved in data processing, from receiving survey data to ordering and receiving full resolution data. Therefore, only brief descriptions of the data products are given. More detailed descriptions of the various data products and processors can be found in a later chapter.

Survey Processing

After the flight season, all HDDCs recorded during the flight season are played back and "surveyed" using the real-time processor available on the flight correlator. The AIRSAR survey product provides an overview of all the data contained on a HDDC and is sent to the PIs to be used for selecting areas to processed at full resolution. The goal is to send all the survey products to the PIs within three months of the end of the flight season. One copy of each of the survey products is also stored at JPL for archival purposes.

The survey photo product shows a single frequency and polarization, usually L-band horizontal transmit and receive (LHH), out of the twelve frequency/polarization combinations normally recorded. The resolution is roughly twenty-four meters in the flight direction and about twelve meters in slant range. From 1990-1991 output was created on a LaserTechniques printer which generated positve transparencies, but since 1992 an Alden thermal printer which produces positive prints directly has been used.

A sample survey photo product is shown in Figure 2.X. The prints show everything on the HDDC, often including calibration data, baseline tests, and other non-image data. The images read from right to left and top to bottom, with radar illumination form the top. The computer-generated label along the top margin shows the data the data were collected, HDDC and part number (1 or 2), followed by the names of the runs contained on the HDDC. For example, the label 93017/3 31 May 93 Mayasquer Cumbal 56-1, 236-1, Azurfal-Galeras 51-1, 231-1, Galeras 360-1, Dona Juana-Huila 32-1 indicates data were acquired on May 31, 1993 on HDDC 93017. The print is the third of the series for this HDDC, which contains two runs at the Mayasquer Cumbal site along flight azimuth 56 and 236 degrees, two runs at the Azurfal-Galeras site along the flight azimuth 51 and 231 degrees, one run at the Galeras site along flight azimuth 360 degrees, and one run at the Dona Juana-Huila site along flight azimuth 32 degrees. The annotation along the far range (lower margin) of each swath indicates the run name, approximate latitude, longitude, and orientation of the aircraft, frequency and polarization, and an important number referred to as the frame count. This six or seven digit number is used to specify data to be processed at full resolution. Numbers are printed every 32,768 frame counts, except in the case of data collected in a special interferometry mode.
 

Full Resolution Processing

Upon receipt of the survey products, the PI selects an area to be processed at full resolution in the frame processing or synoptic mode, fills out an AIRSAR processing request form, and returns it to JPL. (A blank processing request form is included in Appendix X. Samples of completed frame processing and synoptic processing requests are included in Appendix X.) Once at JPL, the request is entered into the processing queue. To distribute data as quickly as possible to as many PIs as possible, the queue is set up in the following manner. The top priority request from each PI is processed first, and once this is achieved, the second priority request from each PI is processed and so forth. Thus, if N PIs submit requests, each PI will receive one full resolution product after N processing runs. Therefore, it is crucial for PIs to indicate the order in which they would like their requests processed. Also, for large experiments involving a large number of investigators, we request a single individual coordinate the processing requests from all involved parties before submitting the requests.

Digital products are delivered on nine-track 6,250 BPI computer tapes or 8 millimeter tapes. JPL can write tapes in the default format of the model EXB-8500 from the Exabyte Corporation or the model EXB-8200, referred to on the Processing Request form as "5 Gigabyte format" and "2 Gigabyte Format," respectively. On the processing request it is necessary that you indicate which format you prefer.
 

Frame Processing

A standard AIRSAR frame product consists of a sixteen-look (where data are acquired with a bandwidth of 20 MHz) or eight-look (where data are acquired with a bandwidth of 40 MHz) "polarization-compressed" digital file for each frequency and a color photo product. A frame processed scene covers 32,768 frame counts. To specify data for processing, decide where the center of the processed frame should lie and interpolate between the printed frame count numbers to obtain the center frame count. This number should be entered on the processing request form.
 

Synoptic Processing

An AIRSAR synoptic product consists of a thirty-two bit VAX floating-point digital image file for three of twelve channels (three frequencies, four polarizations) and a color photo product. In this case, it is necessary to specify both the starting frame count and the three channels to process. On the processing request enter the frame count where the data should begin remembering that a processed run will cover 163,840 frame counts. The choices of frequency and transmited and received polarizations are PHH, PHV, PVH, PVV, LHH, LHV, LVH, LVV, CHH, CHV, CVH, and CVV.
 

Special Processing

Additional special products may be ordered after consultation with JPL personnel, but require departures from the standard processing flow and take more time to complete. The most popular of these is the single-look polarization-compressed complex data, where the four center looks of the sixteen-look data are processed, resulting in four digital files per frequency.
 

Data Distribution

All processed data are distributed by AIRSAR personnel and archived at JPL. To request data that have been processed, contact Michele Vogt.
 

Ancillary Data

Other related sets of information, such as flight logs, in-flight video tapes, and photos taken from the aircraft, are archived at JPL.
 

Data Digests

After the flight season ends, an AIRSAR Data Digest is distributed to PIs and the user community. The purpose of the digest is to provide an overview of that data set along with daily logs of each flight, maps of coverage, and a listing of the raw flight logs generated by the AIRSAR control computer. A limited number of digests are printed and are available from Michele Vogt.