4 th swarm qwg meeting 2 – 5 december 2014gfz potsdam/d on calibrating the magnetometry package...

15
4 th Swarm QWG Meeting 2 – 5 December 2014 GFZ Potsdam/D On Calibrating the Magnetometry Package Data Nils Olsen, DTU Space

Upload: harvey-ramsey

Post on 22-Dec-2015

216 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: 4 th Swarm QWG Meeting 2 – 5 December 2014GFZ Potsdam/D On Calibrating the Magnetometry Package Data Nils Olsen, DTU Space

4th Swarm QWG Meeting 2 – 5 December 2014 GFZ Potsdam/D

On Calibrating the Magnetometry Package Data

Nils Olsen, DTU Space

Page 2: 4 th Swarm QWG Meeting 2 – 5 December 2014GFZ Potsdam/D On Calibrating the Magnetometry Package Data Nils Olsen, DTU Space

4th Swarm QWG Meeting 2 – 5 December 2014 GFZ Potsdam/D

What is a ”Calibration”?

”Calibration” includes two steps:

1. to determine the calibration parameters

2. to apply the estimated parameters to the data

VFM calibration Parameterestimation

Page 3: 4 th Swarm QWG Meeting 2 – 5 December 2014GFZ Potsdam/D On Calibrating the Magnetometry Package Data Nils Olsen, DTU Space

4th Swarm QWG Meeting 2 – 5 December 2014 GFZ Potsdam/D

at

mis

sion

level

at

sing

le-s

ate

llite

le

vel but

co-

est

imati

on

The various calibration steps

VFM calibration(offset, scale-values, non-orthogonalities)

VFM alignment(Euler angles,VFM NEC)

VFM calibration(offset, scale-values, non-orthogonalities)

VFM characterisation(sun-position dependent

disturbance field)

VFM alignment(Euler angles,VFM NEC)

VFM calibration(offset, scale-values, non-orthogonalities)

VFM characterisation(sun-position dependent

disturbance field)

VFM alignment(Euler angles,VFM NEC)

SW-A SW-C SW-B

Euler angles from L2 processing(multi-satellite co-estimation with geomagnetic field model)

VFM characterisation(sun-position dependent

disturbance field)

Para

mete

rest

imati

on

Para

mete

rest

imati

on

Para

mete

rest

imati

on

Page 4: 4 th Swarm QWG Meeting 2 – 5 December 2014GFZ Potsdam/D On Calibrating the Magnetometry Package Data Nils Olsen, DTU Space

4th Swarm QWG Meeting 2 – 5 December 2014 GFZ Potsdam/D

at

mis

sion

level

at

sing

le-s

ate

llite

le

vel but

co-

est

imati

on

Inter-satellite calibration of SW-C

Differential calibration: F(A) F(C)

VFM calibration(offset, scale-values, non-orthogonalities)

VFM alignment(Euler angles,VFM NEC)

VFM calibration(offset, scale-values, non-orthogonalities)

VFM characterisation(sun-position dependent

disturbance field)

VFM alignment(Euler angles,VFM NEC)

VFM calibration(offset, scale-values, non-orthogonalities)

VFM characterisation(sun-position dependent

disturbance field)

VFM alignment(Euler angles,VFM NEC)

SW-A SW-C SW-B

Euler angles from L2 processing(multi-satellite co-estimation with geomagnetic field model)

VFM characterisation(sun-position dependent

disturbance field)

Page 5: 4 th Swarm QWG Meeting 2 – 5 December 2014GFZ Potsdam/D On Calibrating the Magnetometry Package Data Nils Olsen, DTU Space

4th Swarm QWG Meeting 2 – 5 December 2014 GFZ Potsdam/D

Inter-satellite calibration of SW-CHow to calibrate VFM(C) without ASM(C) ?Mapping of F : SW-A SW-C

• FASM(A)

• subtract Fmodel(A), add Fmodel(C) …

• … to obtain an estimate of F’ (C)• use this value to calibrate VFM(C)

all data: s = 0.55 nTnightside non polar: s = 0.28 nT

Comparison FASM (A C) – FASM (C)

daysidenightside

Page 6: 4 th Swarm QWG Meeting 2 – 5 December 2014GFZ Potsdam/D On Calibrating the Magnetometry Package Data Nils Olsen, DTU Space

4th Swarm QWG Meeting 2 – 5 December 2014 GFZ Potsdam/D

Gradient data

• Magnetic field difference between SW-A and SW-C is a key element of Swarm

• Used e.g. for studying small-scale crustal field structures or determination of Field-Aligned Currents (curl-B technique)

• Requires ultra-precise knowledge of the relative calibration/alignment between SW-A and SW-C

• Relative values can be determined with higher accuracy than the difference of absolute values

• Idea: Differential CalibrationDifferential Characterization

Differential Alignment

Page 7: 4 th Swarm QWG Meeting 2 – 5 December 2014GFZ Potsdam/D On Calibrating the Magnetometry Package Data Nils Olsen, DTU Space

4th Swarm QWG Meeting 2 – 5 December 2014 GFZ Potsdam/D

at

mis

sion

level

at

sing

le-s

ate

llite

le

vel but

co-

est

imati

on

Differential Alignment

Differential alignment:

Euler angles VFM(A) VFM(C)

VFM calibration(offset, scale-values, non-orthogonalities)

VFM alignment(Euler angles,VFM NEC)

VFM calibration(offset, scale-values, non-orthogonalities)

VFM characterisation(sun-position dependent

disturbance field)

VFM alignment(Euler angles,VFM NEC)

VFM calibration(offset, scale-values, non-orthogonalities)

VFM characterisation(sun-position dependent

disturbance field)

VFM alignment(Euler angles,VFM NEC)

SW-A SW-C SW-B

Euler angles from L2 processing(multi-satellite co-estimation with geomagnetic field model)

VFM characterisation(sun-position dependent

disturbance field)

Page 8: 4 th Swarm QWG Meeting 2 – 5 December 2014GFZ Potsdam/D On Calibrating the Magnetometry Package Data Nils Olsen, DTU Space

4th Swarm QWG Meeting 2 – 5 December 2014 GFZ Potsdam/D

Differential alignment SW-A vs. SW-C Mapping of B: SW-A SW-C

• BVFM(A) BNEC(A) using STR(A) data and Euler angles for SW-A

• Mapping A C: subtract Bmodel(A), add Bmodel(C) …

• … to obtain an estimate of B’NEC(C)

• B’NEC(C) B’VFM(C) using STR(C) data and Euler angles for SW-C

• Estimate differential Euler angles A C by comparing B’VFM(C) and BVFM(C)

absolute reference level

Eu

ler

angle

a

SW-A SW-C

Difference of (absolute, but less accurate) Euler angles

Differential Euler angle (relative, but more accurate)

aA determined by SW-AaC determined by SW-CaA - aA determined by SW-A – SW-C

How to combine absolute and differential Euler angles ?

Page 9: 4 th Swarm QWG Meeting 2 – 5 December 2014GFZ Potsdam/D On Calibrating the Magnetometry Package Data Nils Olsen, DTU Space

4th Swarm QWG Meeting 2 – 5 December 2014 GFZ Potsdam/D

Difference BVFM(A C) – BVFM(C)

Assumptions:• VFM calibration• VFM characterization• VFM alignment• STR dataare perfectly known (error-free) Non-zero difference can be used to improve • calibration• characterization• alignment

Vector-vector calibration/ characterization allows better y-axis parameter estimation

Page 10: 4 th Swarm QWG Meeting 2 – 5 December 2014GFZ Potsdam/D On Calibrating the Magnetometry Package Data Nils Olsen, DTU Space

4th Swarm QWG Meeting 2 – 5 December 2014 GFZ Potsdam/D

Differential Euler angles between SW-A and SW-C

Page 11: 4 th Swarm QWG Meeting 2 – 5 December 2014GFZ Potsdam/D On Calibrating the Magnetometry Package Data Nils Olsen, DTU Space

4th Swarm QWG Meeting 2 – 5 December 2014 GFZ Potsdam/D

at

mis

sion

level

at

sing

le-s

ate

llite

le

vel but

co-

est

imati

on

Differential Calibration / Alignment

Differential calibration: F(A) F(C)

Differential alignment:

Euler angles VFM(A) VFM(C)

VFM calibration(offset, scale-values, non-orthogonalities)

VFM alignment(Euler angles,VFM NEC)

VFM calibration(offset, scale-values, non-orthogonalities)

VFM characterisation(sun-position dependent

disturbance field)

VFM alignment(Euler angles,VFM NEC)

VFM calibration(offset, scale-values, non-orthogonalities)

VFM characterisation(sun-position dependent

disturbance field)

VFM alignment(Euler angles,VFM NEC)

SW-A SW-C SW-B

Euler angles from L2 processing(multi-satellite co-estimation with geomagnetic field model)

VFM characterisation(sun-position dependent

disturbance field)

Page 12: 4 th Swarm QWG Meeting 2 – 5 December 2014GFZ Potsdam/D On Calibrating the Magnetometry Package Data Nils Olsen, DTU Space

4th Swarm QWG Meeting 2 – 5 December 2014 GFZ Potsdam/D

at

mis

sion

level

at

sing

le-s

ate

llite

le

vel but

co-

est

imati

on

Complete Calibration / Alignment

Differential calibration: F(A) F(C)

Differential alignment:

Euler angles VFM(A) VFM(C)

VFM calibration(offset, scale-values, non-orthogonalities)

VFM alignment(Euler angles,VFM NEC)

VFM calibration(offset, scale-values, non-orthogonalities)

VFM characterisation(sun-position dependent

disturbance field)

VFM alignment(Euler angles,VFM NEC)

VFM calibration(offset, scale-values, non-orthogonalities)

VFM characterisation(sun-position dependent

disturbance field)

VFM alignment(Euler angles,VFM NEC)

SW-A SW-C SW-B

Euler angles from L2 processing(multi-satellite co-estimation with geomagnetic field model)

VFM characterisation(sun-position dependent

disturbance field)

Page 13: 4 th Swarm QWG Meeting 2 – 5 December 2014GFZ Potsdam/D On Calibrating the Magnetometry Package Data Nils Olsen, DTU Space

4th Swarm QWG Meeting 2 – 5 December 2014 GFZ Potsdam/D

Ultimate Goal

• To co-estimate (absolute) calibration/alignment/alignment (CCA) parameters for each satellite, accounting for (better determined) differential CCA values

• Has to be done by co-estimation with geomagnetic field model• Differential CCA parameters determined using more data

(including dayside data, higher geomagnetic activity)• Possibility for vector-vector differential

calibration/characterization• more sensitive to typically weakly determined y-axis calibration

Page 14: 4 th Swarm QWG Meeting 2 – 5 December 2014GFZ Potsdam/D On Calibrating the Magnetometry Package Data Nils Olsen, DTU Space

4th Swarm QWG Meeting 2 – 5 December 2014 GFZ Potsdam/D

Combined absolute and differential CCA:Concept of parameter co-estimation

0

0A A A

C C C

B G m

B G m

/ 2

/ 2A A A

C C C

B G G m

B G G m

/ 2

/ 2

0

A A A

C C C

A C A C

B G Gm

B G Gm

B B G G

/ 2

/ 2A

C

m m m

m m m

A A AB G m

Page 15: 4 th Swarm QWG Meeting 2 – 5 December 2014GFZ Potsdam/D On Calibrating the Magnetometry Package Data Nils Olsen, DTU Space

4th Swarm QWG Meeting 2 – 5 December 2014 GFZ Potsdam/D

Ultimate Goal

• To co-estimate (absolute) calibration/alignment/alignment (CCA) parameters for each satellite, accounting for (better determined) differential CCA values

• Has to be done by co-estimation with geomagnetic field model• Differential CCA parameters determined using more data

(including dayside data, higher geomagnetic activity)• Possibility for vector-vector differential

calibration/characterization• more sensitive to typically weakly determined y-axis calibration

• Estimation of CCA parameters by scientific experts• Application of CCA parameters within L1b processor by PDGS