Transcript Results obtained from 18 months of Swarm satellite data

REPORT ON WP2X00:
TIDAL SIGNAL RECOVERY USING THE
COMPREHENSIVE INVERSION (CI)
RESULTS FROM CHAMP AND SWARM
Nils Olsen, Terence J Sabaka, Lars Tøffner-Clausen
STSE Tides to Sense Earth, MTR
25 January 2016
DTU, Lyngby/DK
Study Work Breakdown Structure
STSE Tides to Sense Earth, MTR
25 January 2016
DTU, Lyngby/DK
Outline
• Extension of CI code to estimate M2 tidal field
• Non-gradient version: CM5
• Including NS (alongtrack) and EW gradients
• Results obtained using CHAMP satellite data
• Published in Sabaka et al 2015 (CM5)
• Some experiments on spherical harmonic truncation level
• Results obtained from 18 months of Swarm satellite data
•
•
•
•
Only field data (no gradient)
Field and NS gradient data (”single satellite solution”)
Field and EW gradient data
Field and NS + EW gradient data
STSE Tides to Sense Earth, MTR
25 January 2016
DTU, Lyngby/DK
Results obtained with CHAMP satellite data
• First successful determination of the spatial structure of the
M2 tidal signal
• Tyler et al (2003) had filter CHAMP data on an orbit-by-orbit basis is order to
be able to extract the M2 tidal signal
• No need for this processing step when using CI
results have been published as CM5 (Sabaka et al 2015)
• M2 tidal field is described by scalar magnetic potential which
is expanded in spherical harmonics up to d/o 36, assuming a
M2 period of 12.42060122 hrs
STSE Tides to Sense Earth, MTR
25 January 2016
DTU, Lyngby/DK
Spatial Power spectrum
STSE Tides to Sense Earth, MTR
25 January 2016
DTU, Lyngby/DK
Maps of |Br| at 430 km altitude
STSE Tides to Sense Earth, MTR
25 January 2016
DTU, Lyngby/DK
Maps of phase(Br) at 430 km altitude
STSE Tides to Sense Earth, MTR
25 January 2016
DTU, Lyngby/DK
Experiments with Different SH Truncation Levels
n = 1 - 18
n = 1 - 36
n = 19 - 36
STSE Tides to Sense Earth, MTR
25 January 2016
DTU, Lyngby/DK
Extension of CI code
Swarm data selection and data use
• Data selection
• 1 Hz data downsampled to 15 sec data
• |dDst/dt| < 3 nT/hr
• Kp ≤30
• East-West/North-South (EW/NS) scalar sums/differences (s/d)
for all local times and all latitudes
• EW/NS vector s/d for all local times and |QD lat| < 55°
• Scalar data nightside for all latitudes
• Vector data nightside for |QD lat| < 55°
• No observatory data
STSE Tides to Sense Earth, MTR
25 January 2016
DTU, Lyngby/DK
Extension of CI code
Model Parameterization (1)
Degree/order (d/o) 90 static (core+crustal) field
d/o 13 linear SV
d/o 1 magnetospheric/induced fields in 1 hr bins
"usual" CI ionospheric parameterization
accounting for 3D induction via Q-matrices
• d/o 36 M2 tidal field (period 12.42060122 hrs)
•
•
•
•
STSE Tides to Sense Earth, MTR
25 January 2016
DTU, Lyngby/DK
Extension of CI code
Model Parameterization (2)
• (Nominal) core determined by all data
except low-latitude dayside
• (Nominal) crust determined by all field data
and all difference data except low-latitude dayside
• Ionosphere determined by all data
• Magnetosphere/induced determined by all data
• (Nominal) M2 determined (mostly) by all difference data
STSE Tides to Sense Earth, MTR
25 January 2016
DTU, Lyngby/DK
Extension of CI code
Major differences from "old" CI model
“Old” CI code: developed and tested during Swarm SCARF development phase
• Direct Huber weighting of s/d rather than formal covariance propagation
from scalar or vector constituents
• Component-only s/d used (like SIFM+) rather than full covariant forms
• The addition of scalar/vector data to s/d data rather than only s/d data
• No vector data for |QD lat| > 55° rather than using vector at all latitudes
(no toroidal field estimated yet)
• Day as well as night data for crust and M2 rather than nightside only
• M2 determined mostly by gradient data
• Core field now varies linearly rather than quadratically
STSE Tides to Sense Earth, MTR
25 January 2016
DTU, Lyngby/DK
Results obtained from 18 months of Swarm
satellite data
Four models have been derived:
•
•
•
•
Only field data (no gradient)
Field and NS gradient data (”single satellite solution”)
Field and EW gradient data
Field and NS + EW gradient data
STSE Tides to Sense Earth, MTR
25 January 2016
DTU, Lyngby/DK
Spatial Power spectrum
STSE Tides to Sense Earth, MTR
25 January 2016
DTU, Lyngby/DK
M2 Tidal Field
|Br| of M2 tide
18 months of Swarm data allows for an
astonishing determination of the M2 tidal signal,
comparable to what has been obtained using
10 years of CHAMP data (at lower altitude)
CI (Swarm)
Sabaka et al., in preparation for GRL
STSE Tides to Sense Earth, MTR
25 January 2016
Model prediction
CM5 (CHAMP)
Sabaka et al., GJI (2015)
DTU, Lyngby/DK
|Br| at 430 km altitude
STSE Tides to Sense Earth, MTR
25 January 2016
DTU, Lyngby/DK
Phase(Br) at 430 km altitude
STSE Tides to Sense Earth, MTR
25 January 2016
DTU, Lyngby/DK
Comparison of M2 Amplitudes and Phases
Br at ground, from CHAMP (CM5, top) and Swarm (CI, bottom)
STSE Tides to Sense Earth, MTR
25 January 2016
DTU, Lyngby/DK
Conclusions
• New version of CI model derived, using only Swarm data and solving for
core, crust, ionosphere, magnetosphere and M2 tidal field
• Major improvement compared to version 01 (determined in May 2015)
• M2 Tidal signal determination from 18 months of Swarm data is
“almost as good as” what is possible from 10 years of CHAMP,
thanks to the novel Swarm constellation gradient concept
• Gradient data is what is responsible for the resolving power of the
oceanic M2 magnetic signal from Swarm data
• EW gradients are doing most of the resolving as compared to the NS
gradients. There seems to be little degradation when NS gradients are
not used
STSE Tides to Sense Earth, MTR
25 January 2016
DTU, Lyngby/DK