Eagle/Hawk/ATM/CASI Boresight calibration

All numbers given as triplets are Pitch-Roll-Heading.

2011 boresight

Period (Julian days)	ATM	CASI	Eagle	Hawk
82 -> end of year			-0.28 0.16 0.38	-0.27 0.27 0.5

2010 boresight

Period (Julian days)	ATM	CASI	Eagle	Hawk
60 -> end of year			-0.18 0.27 0.28	-0.26 0.28 0.5

2009 boresight

Period (Julian days)	ATM	CASI	Eagle	Hawk
(start of year) - 130	-	-	Variable	Variable
131 -> 251	-	-	-0.245 -0.770 0.269	-0.307 0.203 -0.040
252 -> 268	-	-	-0.31 -0.66 0.35	-0.25 0.25 0
268 -> end of year	-	-	-0.32 -0.52 0.26	0.23 0.27 0.0

Notes:

• During the period 001 to 130, there is evidence to suggest that one or both of the sensors was moving and as such, for this period a individual boresight correction is essential for each individual flight line.

2008 boresight

Period (Julian days)	ATM	CASI	Eagle	Hawk
Ethiopia (start of year) - 043	-	-	-0.2 -0.4 1.1	-0.2 0.4 1.9
043 - 116	-	-	-0.20 -0.53 0.6	-0.28 0.2 1
116 -> end of year	0.2 -2.1 0.8	-	-0.2 -0.45 0.5	-0.1 0.2 0.8

Notes:

• Boresight calibration for Eagle and Hawk aligned flight lines to within 10 metres for the majority of lines. A few lines deviated a little further than this but this could well be due to unresolved timing issues. These timing issues would be invisible as the flight lines are fairly straight.
• (adbe) Changed from Ethiopia 0-43, 0-43, and 116-366. Did not seem correct.
• (tipo) Changed 043-116 from -0.25 -0.45 0.6 and -0.2 0.2 1. Did not seem correct.

2007 boresight

Period (Julian days)	ATM	CASI	Eagle	Hawk
(start of year) - 130	0.08 -2.3 -0.75	not mounted	-0.25 -0.5 -1.2	-0.3 0.1 -1.4
166 - (end of year)	0.0 -2.22 -0.75	-0.85 -0.08 -0.4	0.23 -0.28 -1.4	0.0 0.35 -1.4

Notes:

• During the period 130-166, the plane was equipped only with atmospheric measurement instruments for the SOLAS dust cruises. As a result, there are no boresight numbers.
• The 061-130 period was based on flight day 103a, using vector overlays to supplement the lack of cross-overs. This was done because both calibration flights (061 and 130) had severe processing problems.
• The 166 -> end of year period was based on the 166 calibration flight (we also have a 254b towards the end of this period).
• All Eagle and Hawk numbers may have small errors due to timing problems on every flightline. The slippage caused by these errors may result in inaccurate placement of the markers used for boresight calibration.

No 2007 survey of the instrument positioning was done? David Davis said to use the 2006 numbers as the placement of the Applanix IMU and instruments was identical (seems unlikely?). These numbers derived from 2006 batch files and taken from an email from AKW.

• ATM: 0.2655 0.006353 1.741
• CASI: 0.4732 0.007934 1.702
• Eagle: 0.6977 0.00781 1.994
• Hawk: 0.7644 0.00682 2.117

2006 boresight

Period (Julian days)	ATM	CASI	Eagle	Hawk
(start) - (end)	-0.1 -1.95 -0.4	-0.7 0.2 -0.2	0.1 -0.3 -1.25	-0.1 0.3 -1.25

Offset of instruments from GPS antenna (lever arm):

• ATM: 0.2655 0.006353 1.741
• CASI: 0.4732 0.007934 1.702
• Eagle: 0.6977 0.00781 1.994
• Hawk: 0.7644 0.00682 2.117

2005 boresight

Period (Julian days)	ATM	CASI	Eagle	Hawk
?	-0.3 -2.00 -0.2	?	?	?

Offset of instruments from GPS antenna (lever arm):

• ATM: 0.2754 -0.0062 1.7728
• CASI: 0.7633 -0.00373 2.1452

---

Procedure

Copy of email to Andrew Wilson (wikise later)

As I understand it, a calibration flight passes over CEH several times in opposing directions.

The idea is to process the data sets and overlay them, then adjust the PRH numbers until things line up exactly over the calibration points (and hopefully everywhere else).

Because the flights are in opposing directions, one can "focus" the points inwards towards convergence by using estimates of distance and direction of flight/direction to convergence.

Additionally, you have the positions of the calibration points exactly measured (by placing a GPS unit on them for several hours at different times of day) and also have a base station in the flight line (in the CEH tower?).  Once the calibration is correct, the mapped position of these points should match your recorded coordinates (if this is correct, how accurate do you normally get this?).

Obviously, this would need to be done for each sensor (ATM, CASI, Hawk & Eagle).  The LIDAR correction is handled by Cambridge?

Also, which tools do you normally use?  I'm presuming you mostly use Imagine with the GeoTIFFs? 

---

2007 point locations and notes

Calibration zone:

• CEH Monks Wood
• Approximate location:
  • ETRS-89 (WGS-84) XYZ: 3899843.8284m, -16318.7154m, 5030129.9893m
    • Longitude: 0 degrees, 14 minutes, 13.56 seconds (52.474564)
    • Latitude: 52 degrees, 28 minutes, 28.43 seconds (0.237100)
  • OS Grid (OSGB-36): 519857.522, 279414.596
    • Grid ref: TL 52040 88645

Point	ETRS 89 (WGS 84)			Pos Qlty	Hgt Qlty	Pos+Hgt Qlty	OS Grid (OSGB 36)
Id.	X	Y	Z	Quality	Quality	Quality	E	N	H
1	3899990.9122	-16767.0766	5030016.9027	0.0096	0.0146	0.0175	519414.561	279216.784	44.499
2	3899983.7474	-16573.9922	5030023.1466	0.0049	0.0063	0.0080	519607.267	279231.601	44.576
3	3899977.7336	-16320.1818	5030025.6952	0.0120	0.0178	0.0214	519860.745	279244.932	42.279
4	3899843.8284	-16318.7154	5030129.9893	0.0035	0.0069	0.0077	519857.522	279414.596	43.201
5	3899738.4140	-16316.4996	5030210.5789	0.0017	0.0020	0.0027	519856.068	279547.267	42.728
6	3899731.4446	-16091.6668	5030214.0582	0.0002	0.0005	0.0006	520080.550	279561.121	40.668
7 Not shown	3899489.7218	-16085.1161	5030392.4048	0.0040	0.0062	0.0074	520078.781	279861.452	34.480

Please note: point 25 on the previous survey is the same point as point 7 on this survey. As point 7 is an average solution derived from four other points + a temporary base station the result above is a far better solution..

Note that point 7 is not shown on the image below. It actually occurs almost due north of target 6 beyond the Monks Wood buildings at the top right hand corner of the grass square that is actually just off the image. That is the 7th target of slabs used and recorded as the precise GPS points in the originally supplied coords.xls table.

See other attachments below for spreadsheets and vector files.