|
|
AirMISR User's Guide
Contents
This document describes the AirMISR instrument, its applications,
and guides prospective investigators through the steps needed to utilize it.
It provides links to the Multi-angle Imaging Spectro-Radiometer (MISR) program,
the Airborne Science Program, and other supporting organizations.
It is intended for investigators who are already somewhat familiar with
remote sensing techniques, either from aircraft or satellites, and are
interested in multi-angle imagery. The material may also be of use
to engineers, experiment planners, or students.
The Airborne Multi-Angle Imaging Spectrometer (AirMISR) is a
four channel digital camera, mounted on a swiveling drum, that
flies on the high altitude
ER-2 aircraft.
It was developed to support the calibration and validation of the
MISR cameras
on the Terra satellite.
The AirMISR camera uses a MISR brassboard lens and a MISR engineering model focal plane;
giving the airborne sensor spectral and radiometric characteristics very similar to the
satellite instrument.
Since the aircraft flies at about 19 to 20 km altitude, vs approximately
705 km for the Terra satellite, the spatial resolution of AirMISR
is much better than MISR, while the areal coverage is correspondingly reduced.
Example.
MISR and AirMISR are 'pushbroom' sensors, which image a cross-track slit beneath them during
each scan. A two dimensional scene is built up by successive scans as the
spacecraft or aircraft fly over the scene.
While MISR has nine cameras which image the earth continuously,
AirMISR's single camera is swiveled to successive
look angles
to complete a desired set of multi-angle images.
This means that AirMISR, unlike MISR, only produces a complete multi-angle image set
over a single patch in the middle of the imaging run.
Performance Specifications
Camera Specifications
| Number of active pixels | 1504 per band |
|---|
Cross track field of view
(swath width) | 29.8° |
|---|
| Along track instantaneous field of view (IFOV) | 0.3 milli-radians |
|---|
| Spectral bands: | 446.4 nm (blue), 557.5 nm (green),
671.7 nm (red), 866.4 nm (near-IR), |
|---|
| Bandwidths: | 41.9 nm, 28.6 nm, 21.9 nm, 39.7 nm |
|---|
| Out-of-band/in-band ratio: | 0.8%, 2.3%, 4.9%, 1.8% |
|---|
| Digitization: | 14 bits |
|---|
Radiometric calibration
absolute uncertainty: | 5% (1 standard deviation), at maximum signal |
|---|
| Signal to Noise Ratio (SNR) | > 900, excluding anomalous pixels |
|---|
| Viewing angles | 9, identical to MISR;
±70.5°, ±60°, ±45.6°, ±26.1°, and 0° (nadir) |
|---|
Platform/Instrument Specifications
based on nominal ER-2 operating parameters
| Flight altitude | 20.5 km above mean sea level (MSL) |
|---|
| Ground speed | 204 meters/second |
|---|
| Duration of imaging run | 12 minutes |
|---|
| Length of imaging run | 148 km |
|---|
| Maximum number of runs per flight | >14 |
|---|
| Maximum flight duration | >6.5 hours |
|---|
The spatial resolution and extent of each AirMISR scene depends upon the aircraft
altitude and target elevation. For a target at sea level,
the following table (&
figure, 7kb)
shows the coverage of each of the nine images acquired in a run.
| | Camera Angle |
Df,Da,
±70.5° |
Cf,Ca,
±60° |
Bf,Ba,
±45.6° |
Af,Aa,
±26.1° |
An, nadir,
0° |
|---|
| Swath Width (km) | 32.7 | 21.8 | 15.6 | 12.1 | 10.9 |
|---|
| Along track image length (km) | 27.0 | 17.0 | 11.3 | 9.6 | 9.6 |
|---|
| Cross track pixel size (m) | 21.7 | 14.5 | 10.4 | 8.1 | 7.3 |
|---|
| Along track pixel size (m) | 55.8 | 24.9 | 12.7 | 7.7 | 6.2 |
|---|
Line spacing (distance
traveled between scans) | 8.3 m, for all angles |
|---|
| %Overlap/Underlap (gap) | 85% | 67% | 35% | -8% | -34% |
|---|
To simplify requirements on the end user, and to maintain commonality with
MISR algorithms, the standard AirMISR data processing normalizes all images to
27.5 meter resolution, 10% of the MISR resolution.
These images are also geo-rectified, with North at the top, and
the effects of aircraft motion are removed to the extent possible.
AirMISR Data Quality
Each of the AirMISR data sets available from the Atmospheric Sciences Data
Center (ASDC) are listed at include product-specific data quality statements.
Most of these statements can be represented by the following generalized
summary statements.
Radiometric Accuracy
The radiometric calibration of AirMISR has been done using the same procedures
as used to calibrate the MISR cameras. Cross-comparison studies with MISR,
Landsat and Vicarious Calibration studies indicate AirMISR has a 5% bias as
compared to other sensors, reporting higher radiances.
Vertical striping may at times be viewed in AirMISR imagery. These are
insignificant in terms of radiance variability, as they are typically
on the order of 0.5%.
One difference between MISR and AirMISR radiometric performance is the
camera-relative uncertainty, which is smaller for AirMISR in that this
aircraft sensor consists of only one gimballed camera. Thus, errors at
one view angle, relative to another are not present. This is not the
case for MISR, which makes use of nine cameras to acquire the mult-angle
data.
The values quoted for the systematic component of the radiometric
uncertainty, based on vicarious calibration of the instrument,
in fractional units, are
- abs_sys_error = 0.050
- cam_sys_error = 0.000
- band_sys_error = 0.010
- pixel_sys_error = 0.005
These systematic components are combined with the signal-to-noise
ratio (SNR), to determine the total error uncertainties. As SNR is
signal dependent, the uncertainties are likewise signal dependent.
The SNR, at two radiance input levels, are as follows
- SNR(at an equivalent reflectance of 1.0) ~ 1000
- SNR(at an equivalent reflectance of 0.05) ~ 200
Using these, the total radiometric uncertainties can be determined
- abs_total_error=sqrt(abs_sys_error2+(1/SNR)2)
- cam_total_error=sqrt(2)/SNR
- band_total_error=sqrt(2)*sqrt(band_sys_error2+(1/SNR)2)
- pixel_total_error=sqrt(2)*sqrt(pixel_sys_error2+(1/SNR)2)
Ghost image effects
The MISR and AirMISR camera designs share a common heritage,
hence performance evaluation studies conducted for one are
also applicable to the other. The radiometric accuracy for the
MISR cameras are described in Bruegge, et al. (2002).
One such study, involving AirMISR and MISR data sets acquired
over the Chesapeake Lighthouse, August 2001, showed signal-dependent
radiance ratios. This has been attributed to ghost-image effects,
in which light from the brighter land is reflected into the darker
ocean portions of the scene. The darker targets have proportionally
greater radiometric error due to these effects.
AirMISR is less sensitive to such errors, in that scene uniformity
over the 10 km swath is typically greater than observed for the
380 km MISR swath widths. Thus background contamination of the
signal is less for AirMISR.
AirMISR Geometric Data Quality
MISR geolocated and co-registered radiance data sets
(Level 1B2 products) can be produced by initial, manual,
and automated algorithms. The manual process has been
used for the majority of products delivered to date.
For the manual L1B2 product, the geometric calibration has been
performed prior to orthorectification to the UTM map projection
grid. A set of ground control points collected from US 124000
topographic maps are used to improve geometric data quality.
These points are identified in AirMISR imagery in order to remove
static errors in the camera pointing and airplane position.
Only runs taken over clear land are used for ground control point
identification. It is expected that geolocation errors of about
1000 meters for nadir view to up to 6000 meters for the most
oblique view (for the initial L1B2 product) are reduced down to
an average of about 150 meters regarding both absolute geolocation
and coregistration between nine view images. The remaining errors
can be regarded as a result of the dynamic errors in airplane
attitude and position which are not modeled in the current
calibration algorithm.
AirMISR has been employed to radiometrically calibrate the MISR cameras,
to measure directional reflectance characteristics of soils, vegetation,
snow, and clouds, and to measure the amount and type of aerosols in the atmosphere.
AirMISR can serve as a surrogate MISR, revisiting a scene more frequently than the Terra
satellite's 16 day repeat cycle, or on days during focused investigations in which weather
conditions are ideal but no satellite overpass are scheduled.
AirMISR may be used to explore other sun angles, azimuths, or timing than those possible
with MISR.
The instrument may also be flown with other airborne instruments, such as
sounders,
profiling LIDARs,
hyperspectral scanners,
or
other EOS instrument simulators,
etc., to provide a comprehensive dataset on a single scene.
Here are a few
published papers that have used AirMISR data.
The Jet Propulsion Laboratory (JPL)
operates and maintains the AirMISR instrument for NASA, and and processes the
data into georectified radiance products as part of the MISR data processing task.
JPL is responsible for scheduling the instrument, staffing field campaigns,
repairing, maintaining, calibrating, and upgrading the instrument,
and delivering the final data products to the investigator and to the
Langley Atmospheric Sciences Data Center
DAAC,
where they are made available publicly.
Most of the costs of this activity are borne by the EOS/MISR program, as
part of the calibration & validation effort for the satellite instrument.
Some excessive or unusual costs, such as deployments or major instrument
changes, may require additional funding by an investigator or program.
JPL can also assist in identifying MISR overpasses and existing datasets,
and has some
ground instrumentation
that may be of interest to investigators.
NASA's Dryden Flight Research Center
operates the ER-2 aircraft, and manages the process of soliciting and reviewing
flight requests, in conjunction with NASA Headquarters.
Dryden will schedule the use of the aircraft to maximize the science return among
multiple instruments or users and within budget and staffing constraints.
Dryden provides the aircraft pilots, maintenance crew,
and mission managers who plan and execute a campaign.
The investigator is responsible for:
- Designing the experiment which requires the unique assets of AirMISR and the ER-2
(along with any other instruments).
- Submitting a flight request (with help from JPL, if desired) to NASA Dryden.
- Planning the experiment, which includes specifying target coordinates, flight line orientation,
timing constraints, weather requirements, etc.
- Funding any flight hour charge or mission peculiar costs not borne by the MISR program
(to be negotiated).
- Answering questions from the pilots and mission planners to resolve ambiguities or conflicts.
- Coordinating with ground teams or other groups that are essential to mission success.
- During multiple sensor field campaigns, working with other investigators to reach consensus
on flight plans involving multiple, perhaps conflicting, objectives.
Daily participation at the aircraft deployment site may be required.
- Reviewing data acquisition logs to ensure the images acquired are satisfactory,
as soon as possible after each flight.
- Receiving and checking the processed data when it is delivered.
- Providing feedback to the JPL and Dryden team on the conduct and success of the experiment.
An experiment summary report is requested as soon as possible after the campaign.
Examples of such reports are on the MISR calibration/validation group's
web page.
- Recognize JPL and the MISR project in any publication using AirMISR data.
Informal queries may be directed to JPL at any time to determine technical feasibility
and scheduling possibilities for new AirMISR experiments.
This is recommended before beginning the formal flight request process, to ensure
instrument suitability for the measurement proposed and availability in the time
frame required.
The formal process of requesting time on the ER-2 aircraft, with AirMISR, begins
with a flight request to NASA Dryden. This request captures information about the
objectives of the proposed mission, relevance to NASA goals and programs, and
as many specific details of the experiment that are known at the time.
The annual cycle of solicitation flight requests for the next fiscal year normally begins in May.
The requests are usually due back at Dryden sometime in early June.
Dryden uses these requests to assess where and when to deploy the ER-2 aircraft to most
efficiently satisfy as many requests as possible.
Each flight request is assessed for feasibility, scheduling constraints, and for cost,
in terms of the number of aircraft flight hours required.
This information is then provided to NASA headquarters for review.
NASA's Office of Earth Science may approve the request, deny it, or approve it with a reduced
number of flight hours.
Usually, the decision on which flight requests are approved is known by late September or October
(the beginning of the fiscal year). Requests may be sent in at other times of
the fiscal year, but these may not be able to influence the deployment schedule.
The forms, and additional information on the process, are available at the Suborbital Science Program
Web Page.
After a request is approved, Dryden and JPL will coordinate to schedule
a deployment of AirMISR to wherever the ER-2 is to be based.
Communication with the investigator will continue as needed
during the planning and execution stages.
To adequately specify a desired AirMISR data acquisition, the
investigator must provide information about the location of the scene
or point to be imaged, the desired weather and sun illumination,
the timing constraints important to the experiment, and any constraints
relative to other instruments, either on satellites, other aircraft,
or on the ground. This information is used to plan and execute the
flight. The following sections detail the information needed in advance,
and give a few examples.
Target Coordinates
The center point of AirMISR image acquisition should be specified,
and the pilots and mission planners will determine the correct
beginning and end points for the aircraft imaging run.
Latitude and Longitude are best given in degrees and decimal minutes,
to a tenth of a minute, if possible.
If another set of units are used
(e.g., decimal degrees, or degrees, minutes, and seconds),
this should be clearly shown.
The standard datum for ER-2 flight planning is WGS-84.
If another datum is used to specify target coordinates
(NAD-27, et.al.), this should also be clearly stated.
A complete mission may have more than one center point or target,
and more than one run over a target. If multiple runs are requested,
the investigator should specify the timing priorities or relative order
desired for each target and line.
Given the finite speed of the aircraft, a set
of four to six runs over one or two targets may require more than
an hour to complete. If one run has a critical timing constraint
(e.g., coincidence with a satellite pass, or a specific sun angle),
the investigator should clearly indicate which order or timing
they prefer.
Target Elevation
The ER-2 aircraft flies at a near constant altitude, and all
AirMISR imaging runs have been standardized to 12 minutes.
This means that the actual look angles will vary slightly
with different target elevations. This can be corrected for, if
the target elevation above mean sea level is specified.
Elevation need only be specified to the nearest 500 meters,
if it is not precisely known or if it varies over the scene.
If other units are used (e.g., feet above mean sea level),
this should be clearly noted.
If the target elevation is not known (e.g., for clouds or
other atmospheric phenomena),
the best available guess must be provided to the field team
at least four hours prior to the planned take-off time.
Up to two different target elevations can be pre-programmed
into the instrument before flight. At least three hours prior
to takeoff, the pilot must be given clear instructions for
which of the two settings is to be used for each target.
Line Headings or Azimuths
Besides the center point and length (fixed at 12 minutes),
the imaging run is defined by its direction over the target.
This should be specified in degrees from true North.
If the azimuth is a function of time, sun angle, or dependent
on a satellite orbital geometry, the investigator must
work with the field team to ensure that the resulting headings
are calculated correctly.
Geo-location Imaging Runs
As mentioned in the performance section, the geo-location accuracy of
AirMISR imagery can be improved from 6 km at the most oblique views to
150 m at all angles by identification of ground control points within
the imagery.
Part of this improvement is obtained by modeling the the angular offsets between
the camera and the aircraft's attitude frame of reference. These angles can be
indirectly measured by comparing two sets of images taken of
the same scene on the same flight from two reciprocal (±180°) headings.
A minimum of two reciprocal lines are required in order to separate static
camera pointing errors from static aircraft position errors.
The value to the investigator of these 'geo-location' runs is usually
considered worth the additional flight time (~30 minutes, or less).
These imaging runs should be cloud-free, and contain some pronounced features,
such as roads, riverbeds, fence lines, or major buildings.
For investigators whose target of interest is a ground point, two passes
over the center point at opposite headings should suffice.
For other investigators studying diffuse, time dependent, or low contrast targets
(such as open water, clouds, snow fields without plowed roads, etc.), another ground
point should be chosen enroute between the airfield and the primary target.
Weather Constraints
The investigator should specify the weather conditions
(cloud cover, aerosol loading, sea state, etc.) that
are required for the experiment.
For many surface observations, clear skies are preferred.
However, the investigator should
carefully assess what is the maximum amount of cloud cover
that would be tolerable in this case. If the scheduled calendar
window for the experiment is drawing to a close, and the desired
weather conditions have not materialized, the investigator
may wish to consider relaxing the constraints in order to
increase the odds of a data acquisition.
The investigator is welcome to participate in the
weather assessment and decision each day of the campaign.
If clear skies are the requirement, the ER-2 operations crew may
be able to assess the conditions for the investigator, as this
is a common request for remote sensing investigations.
For any weather constraint other than clear skies, the investigator
is requiredto participate in each day's assessment
or provide the information needed to make the proper decision.
The ER-2 pilots will assess and judge whether the weather
at the deployment site will permit safe operation of the aircraft.
This decision takes precedence over target weather conditions.
It should be apparent that the decision on whether and where to
fly can only be made on the basis of forecast information and current
weather observations. Sometimes, forecasts are wrong or observations
are inaccurate, and a good opportunity is missed or,
the data acquired does not meet the experiment objectives.
This is part of the challenge of field work, and can only
be partially mitigated by:
- Scheduling flexibility: fewer constraints = more possibilities.
- Meteorological forecast support during the campaign.
- Timely observations from the target site, perhaps from a ground observer.
- Budget, calendar, and flight hour margin: more hours allocated to an
experiment allows more attempts to be made, or re-flight(s) to be made.
Many investigations using AirMISR rely upon coincident satellite
observations. The usual objective is to match AirMISR's
image acquisition to the satellite instrument's as closely as possible,
in spatial coordinates, viewing geometry, and timing. Some compromises
must be made, due to the differences in altitude and speed.
If an investigation uses a fixed target, the days at which the satellite
views the scene can be predicted in advance if the orbital geometry is
stable. Upon request, a member of the MISR data processing staff can run
JPL-resident software specialized to predict overpass information for the
MISR instrument on board the Terra satellite. This tool provides dates and
times of overpasses, as well as viewing geometry and solar angles.
Output from this program includes
plots (48 kb) and
tabular summaries.
When using this or any other tool, it is recommended that the latest orbital
elements be used, the resulting overpass times be expressed in UTC,
and ground coordinates be converted from geocentric to geodetic latitudes
before being used to specify aircraft flight paths.
The timing tolerance between the satellite and aircraft acquisitions should
be clearly specified. For many investigations in which surface or atmospheric
conditions can change rapidly, AirMISR's imaging runs should be within minutes of
the satellite pass.
Other investigations, such as geology, may be able to accept longer intervals,
up to days, between the satellite overpass and the aircraft flight, as long as the
differences in atmospheric conditions and illumination can be accounted for.
If the investigator specifies a timing requirement relative to an
overpass time, this will take precedence over any sun angle specified
for the mission.
To match the viewing geometry, the aircraft normally flies an imaging
run parallel to the satellite ground track, centered over the target.
If the cross track viewing angle from the satellite to the target is large,
it may be desired to offset the aircraft's track by an amount that produces
an equivalent cross-track viewing angle.
The amount and direction of this parallel offset should be specified
at the same time the ground track is provided for flight planning, so that the
resulting plan correctly identifies any potential airspace conflicts.
The investigator should notify ER-2 operations or the instrument team
during the campaign if a malfunction of the satellite or its instrument
prevents a meaningful aircraft mission.
The investigator is responsible for ordering the data acquisition by
the satellite instrument, including any scheduling changes required.
Surface measurements are often used in conjunction with airborne and spaceborne
remote sensing missions.
Some experiments use automated devices, but most require field personnel to
travel to the site of interest, deploy specialized equipment, and operate them
during the course of the campaign.
These activities normally have expenses and a schedule window of their own,
which may constrain the aircraft schedule.
It may be desired to cancel AirMISR missions during an interval in which
ground teams are not available or when a key instrument is not functioning properly.
Alternately, for some experiments the key instrument is AirMISR, and ground
instrumentation is of secondary importance.
The ground-rules for such decisions should be agreed upon with the investigator
before the campaign begins, and updates provided as necessary during the
campaign.
If the ground team includes the investigator, they may be able to provide
a local assessment of weather or surface conditions on the day of each attempted
flight. This requires reliable communications and a coordinated schedule of
decision making. Best practice is to confirm on the day prior to a mission that
all required elements are in place, and to arrange a time for a telephone call
between the field team, the aircraft operations crew, and the investigator
on the morning of each attempt.
A list of ground instruments used by JPL in conjunction with previous
AirMISR campaigns can be found
here.
Examples
Below are two example mission specifications.
The first is a repeated pattern of straight lines over two targets,
for vicarious calibration of the MISR cameras.
The second is a pattern of four lines at differing azimuths over a
snow-field for BDRF measurements.
Example: Lunar Lake & Railroad Valley Flight Plan
The objective of the 2001 vicarious calibration campaign was to update the radiometric
calibration of the MISR cameras. The approach used was to
view a bright, uniform target (a dry lakebed) with AirMISR simultaneously with a MISR
acquisition and surface observations.
The flight lines were oriented to match the satellite ground track,
in order to provide the same viewing geometry.
Two dry lakebeds were used: Lunar Lake has higher reflectance and is very uniform;
Railroad Valley was the site of University of Arizona ground teams supporting MODIS
and ASTER calibration.
Map, 179kb .
Multiple runs were requested over each target, to maximize the chances of acquiring
a complete set of nine images. The runs began before, and ended after, the MISR overpass.
Example AirMISR Mission Specification
| Center points/target coordinates and elevation |
| Latitude | Longitude | Elevation,
MSL |
|---|
| Lunar Lake | 38° 23.8'N | 115° 58.9'W | 1730 m |
| Railroad Valley | 38° 29.8'N | 115° 41.5'W | 1430 m |
| Flight line orientation, targets, and timing |
| Run Number | Target | Heading | Timing |
|---|
| 1 | Railroad Valley | 13° | |
| 2 | " | 193° | |
| 3 | Lunar Lake | 13° | on center at 18:52 UTC |
| 4 | " | 193° | |
Weather and other Constraints |
| Cloud cover | <5% |
| Surface condition | No blowing dust, as assessed by ground team |
| Coincident with satellite |
Yes, Terra. Flights only requested on days with acceptable overpasses;
list of dates to be supplied. |
| Sun angles |
Superseded by satellite overpass timing requirements |
| Campaign window |
May 20 to July 7, 2001 (to be coordinated with ground teams) |
|
Results
The preceding requirements were used to plan the flight of June 30, 2001.
The results of this mission are detailed
here.
The timing of the AirMISR imaging runs, the MISR overpass, and the
movement of the sun during the runs, is shown in
this figure (8 kb),
while the flight track of the aircraft is
here (10 kb).
Example: "Star Pattern" Flight Plan
The second example is of a single target site imaged at multiple azimuths.
The objective was to validate an existing scheme for converting
measurements of snow bidirectional reflectance to snow albedo.
The target was a uniform snow field near Steamboat Springs, Colorado.
Map, 168 kb.
Four AirMISR data runs were requested at three headings;
a pair up and back along the Terra track,
another along the solar principal plane,
and one perpendicular to the solar principal plane.
Data was to be acquired simultaneously with surface measurements and a MISR
overpass under cloud-free skies.
Example AirMISR Mission Specification
| Center points/target coordinates and elevation |
| | Latitude | Longitude | Elevation,
MSL |
|---|
| Steamboat Springs |
40° 26.0'N |
106° 49.0'W |
2050 m |
| Flight line orientation and timing |
| Run Number | Requirement | Derived Heading | Timing |
|---|
| 1 | Along Terra track | 193° | on center at 18:12 UTC |
| 2 | To sun | 160° | |
| 3 | Perpendicular to line 2 | 250° |
| 4 | along Terra track, reverse of line 1 | 13° | |
Weather and other Constraints |
| Cloud cover | <10%, no cirrus |
| Surface condition | Snow covered; as assessed by PI |
| Coincident with satellite |
Yes, Terra. Flights only requested on days with acceptable overpasses;
list of dates to be supplied. |
| Sun angles |
Superseded by satellite overpass timing requirements |
| Campaign window |
February 3 to March 7, 2001 (to be coordinated with ground teams) |
|
Results
The preceding instructions were used for a campaign in the early Spring of 2001.
During the six-week campaign window, no clear days coincided with a MISR overpass.
On March 8, the weather cleared, and AirMISR imagery was acquired simultaneously
with surface measurements. The results are detailed
here.
Here is the aircraft's flight track (10 kb)
and the timing of the runs (7 kb)
Changes to Flight Plans
It is understood that experiment plans may change from the
time the original flight request is submitted to the time
the first flight is attempted. The instrument and aircraft crews
will work with the investigator to accommodate as many changes
as possible. However, there are some constraints and recommendations
the investigator should be aware of.
Changes to the flight lines may add additional time to the flight;
this must be compared with the flight hours allotted for the
original flight request.
NASA's Office of Earth Science reviews and approves flight requests
for specific purposes. Radically changing the objectives or approach
of the experiment after the request is approved may require additional
justification.
Changes in flight line coordinates or headings could require the
aircraft to enter an area of restricted airspace that was previously
not an issue. Clearance to enter this airspace may take days or weeks
to be granted, and may require fees to be paid or the approval of
written agreements.
If permission is granted, it may come with additional scheduling
constraints to the campaign (i.e., not between certain times of day,
or not on a particular day). If possible, the investigator should
allow at least a day to re-plan the new flight lines and
discover these issues in time to resolve them. The changes may also
prompt feedback from the pilots which the investigator must respond to,
such as a suggested change in routing that removes a conflict, but may
impact the experiment. This feedback and interchange can be done by
phone or e-mail, if the investigator is not co-located at the airfield
with AirMISR and the ER-2.
Flights over foreign airspace, including Canada and Mexico, require
additional advance clearance requests. NASA requires that foreign
governments review and approve any overflight by NASA-owned aircraft.
This involves sending advance information about the nature of the mission
to NASA headquarters, which informs the Department of State, other US agencies,
and the foreign government's embassy. While permission may be granted
for the overflight, weeks or months may be needed for the dialogue and
negotiations to conclude.
During the flight, very little can be done to change the mission.
If weather conditions or other variables change enough to render the
mission useless, it may be possible to contact the pilot by radio or
satellite phone to abort the mission. This contact will be made only
through the ER-2 operations crew at the basing airfield. Successful
communication cannot be guaranteed.
After the flight, a preliminary examination of the images obtained is
performed in the field. This is intended to verify that the camera
operated properly, to assess whether the target was acquired at all
nine view angles, and determine the presence or abscence of clouds in
the scene. This field inspection does not change any of the data recorded.
The original raw image files, along with navigation data and engineering
data recorded in flight, are sent back to the MISR Science Computing
Facility (SCF) at JPL for processing and distribution.
The processing results in a set of radiometrically corrected and
georectified images, resampled to a uniform resolution, stored in
Heirarchical Data Format (HDF) files.
These are then delivered to the investigator on removable media.
Some subset of the data may be distributed by FTP.
A copy of the data is also sent to the NASA Langley
Atmospheric Sciences Data Center (Distributed Active Archive Center), where it is permanently archived and made
available for other researchers.
The steps involved in this processing are:
- Resampling the recorded navigation data, including aircraft position and attitude,
to a uniform 64 Hz time resolution.
- Converting the raw digital numbers stored in the image files into radiometric
values by applying the latest calibration coefficients for each pixel in the
AirMISR camera. The outcome of this is termed a "Level 1B1" product,
which is spectral radiances stored in HDF format. Browse images are also produced
at the full resolution of the AirMISR camera, but not corrected for aircraft
motion.
- The images are then projected onto a terrain model, and resampled to a uniform
27.5 meter resolution on a Universal Transverse Mercator (UTM) projection.
This involves both automated routines and manual
identification of tie-points from LandSat imagery that can improve the
geo-rectification. These images are "Level 1B2" products.
- Additional processing can be performed to locate target points within
the images, to produce browse images or RGB composites, or to convert
the images to ERDAS Imagine® or other formats.
The time required for processing depends on the amount of data acquired,
other demands on the SCF due to MISR operations, and the quality of the
recorded aircraft navigation data.
Normal practice is to deliver initial Level 1B1 products within six weeks of
the flight, and final Level 1B2 georectified images within six months or less.
AirMISR has operated since 1998, for various campaigns across the continental US,
Alaska, and in South Africa in 2000.
Some of the early data acquisitions were affected by various instrument problems
that were fixed in later deployments. The highest quality AirMISR datasets have been
sent to the
Atmospheric Sciences Data Center for archival and distribution. The ASDC provides browse images, a description of the
data quality, and instructions on how to order these datasets.
- Email AskMISR
Navigation:
Abstract|
Overview|
Applications|
Orgs & Resps|
New Data Req |
Flight Line Planning |
Data Processing |
Existing Data |
Contacts|
Links
|
