1 of 21

The Electrojet Zeeman Imaging Explorer mission design through OSSEs��

Rafael L.A. Mesquita¹, Jeng-Hwa (Sam) Yee¹, Jesper W. Gjerloev¹, Nelli Mosavi¹, Viacheslav G. Merkin¹, Astrid Maute², Kareem A. Sorathia¹, Karl M. Laundal³, and Michael Madelaire³.

⁽¹⁾The Johns Hopkins University Applied Physics Laboratory, Laurel, MD.

⁽²⁾High Altitude Observatory, National Center for Atmospheric Research, Boulder, CO.

⁽³⁾ University of Bergen, Bergen, Norway.

Rafael Mesquita (rafael.mesquita@jhuapl.edu)

2 of 21

Observing System Simulation Experiment (OSSE)

2

What is an OSSE?

An OSSE is an exercise to simulate what an instrument would realistically measure if it sampled a modeled atmosphere. It is not just to sample the model, but simulate the instrument and its measurement capabilities with noise and errors.
In the case of EZIE, it consists in simulating the satellite measurement, if it flew through a realistic modeled atmosphere to optimize mission design and operations, establish system requirements, test and ready retrieval algorithms, demonstrate measurement capabilities, ensure science closure, etc.

EZIE measurement forward model using realistic atmosphere

SSWRF II

03 - 07 October 2022

Rafael Mesquita (rafael.mesquita@jhuapl.edu)

3 of 21

Why is it important to have an OSSE?

The OSSE is the tool that allows the team to define the mission design and the necessary parameters to achieve closure of scientific goals:

Number of spacecraft and flight configuration;
Separation between look angles;
Cadence of measurements;
SNR for closure of scientific goals;
Spacecraft jitter;
Precision and accuracy;
…

3

SSWRF II

03 - 07 October 2022

Rafael Mesquita (rafael.mesquita@jhuapl.edu)

4 of 21

OSSE – GAMERA MHD Modeling

4

Rafael Mesquita (rafael.mesquita@jhuapl.edu)

Magnetosphere from global MHD GAMERA model

Auroral electrojet currents calculated with GAMERA model output magnetic fields.

We can then fly the satellite through the model at any configuration (different MLT rotations).

0 MLT

18

6

0.1 µA/m²R_E

0.00

0.01

0.02

0.03

0.04

0.05

0.06

0.07

0.08

0.09

0.10

J_HALL

(µA/m²R_E)

SSWRF II

03 - 07 October 2022

Rafael Mesquita (rafael.mesquita@jhuapl.edu)

5 of 21

OSSE – Basic Steps

5

Rafael Mesquita (rafael.mesquita@jhuapl.edu)

Simulated Tυ, Beam 3

Tυ sensitivity to 1 nT change in B_z

Simulated Retrievals of Vector B

One simulated observation every 2.0 seconds using GAMERA, 98 kHz per channel, and 128 channels in the simulation.

Sensitivity plot that relates brightness temperature to magnetic fields of the simulated measurements.

Retrieved magnetic fields (scatter plot) with radiance noises of 1.45 K in comparison with GAMERA at 80 km.

(K/nT)

SSWRF II

03 - 07 October 2022

Rafael Mesquita (rafael.mesquita@jhuapl.edu)

6 of 21

OSSE – Basic Info

6

DOI: 10.1002/9781119815631.ch21

B magnitude

Peak separation

6

4

8 x 10⁴

2

0

-2

8 x 10⁴

6

4

2

0

-2

Frequency (MHz)

|B|

B

Uncalibrated T_B (K)

-4

-1

-0.5

0

0.5

1

SSWRF II

03 - 07 October 2022

Rafael Mesquita (rafael.mesquita@jhuapl.edu)

7 of 21

OSSE – Radiance Simulator Top Level Diagram

7

Populates the points along the line-of-sight from the ground to a hard-coded altitude. The atmospheric parameters can be populated either from file (WACCM-X or TIEGCM) or calculated using MSIS and dBs come from either GAMERA (high-latitude) or WACCM-X (low-latitude enabled with python).

Read atmosphere state

Populate the line-of-sight atmos

Read mag. perturbation (dB)

Populate the line-of-sight dBs

Populate the line-of-sight with IGRF

Unit conversions are done post spectra calculation.

Calculate O2 Zeeman spectra

Calculate O2 Zeeman emission propagation matrix

Populate line-of-sight effect of spacecraft motion and Earth rotation

Output are IDL Structures:

obs: contents of the configuration file.
pos: spacecraft position.
look: pointing directions of each beam.
ezie: what the OSSE is using to calculate the splitting
spec: spectra (brightness temperature vs frequency of the spectropolarimeter vs positions/time

Partially python enabled depending on version of OSSE

Done exclusively in IDL

Read simulated orbit from file, adjust orbit based on configuration files, and output structure “los”

Read configuration files

Establish spacecraft path

Establish sub-limb geometry

Read HITRAN and est. frequency domain

*Within blocks, tasks run in series from top to bottom and each block from left to right.

**All quantities are delivered in geomagnetic and geographic coordinates

***The above diagram is repeated with and without dB

SSWRF II

03 - 07 October 2022

Rafael Mesquita (rafael.mesquita@jhuapl.edu)

8 of 21

OSSE – Jacobian Calculator Top Level Diagram

8

“Read and populate” LOS IGRF and atmos are similar to radiance sim. Motion related issues depend on the OSSE run.

Radiance simulated file features config info. Establish sub-limb geometry could potentially be saved in radiance sim and excluded from Jacobian.

Read HITRAN and est. frequency domain

Establish sub-limb geometry

Read sim. radiances w/o current

Read and populate the line-of-sight with IGRF

Read and populate the line-of-sight atmos

Change in magnetic fields are to estimate the derivatives. These changes are done for each component of the magnetic field (Bn, Be, Bd). Unit conversions are done post spectra calculation.

”change_mag_fields” to add 10 nT to the IGRF magnetic field

Calculate O2 Zeeman emission propagation matrix (with charge)

Calculate O2 Zeeman spectra (with change)

Output are IDL Structures:

obs: contents of the configuration file.
pos: spacecraft position.
look: pointing directions of each beam.
ezie: what the OSSE is using to calculate the splitting
sens: sensitivity (brightness temperature per nT vs frequency of the spectropolarimeter vs positions/time).

Populate spacecraft motion and Earth rotation

Done exclusively in IDL

SSWRF II

03 - 07 October 2022

Rafael Mesquita (rafael.mesquita@jhuapl.edu)

9 of 21

OSSE – Retrieval Top Level (Precision Calculator)

9

Some of the basic parameters defined upfront could be calculated from the input files.

Define basic parameters:

Number of polarizations;
Error level (Tsys)
Look and observation cases;
Wind and shift presences;
MLT rotation;
Etc.

Read input file:

Radiance simulator with no current/dB (spec0)

Do the MEM instrument spectral radiance convolution of spec0 (output stokes0)

Output are IDL Structures:

obs: contents of the configuration file.
pos: spacecraft position.
look: pointing directions of each beam.
ezie: what the OSSE is using to calculate the splitting
sensor: sensor information.
sens: sensitivity (brightness temperature per nT vs frequency of the spectropolarimeter vs positions/time
rad_no_current, rad_with_current_no_noise, rad_with_current_with_noise: radiances without current, radiance with current and no noise, radiance with current and with noise.
mag_current: output measurements.
mag_current: model magnetic field.

At this point we have stokes0, stokes, and the sensitivity for the dB components.

Read input file:

Jacobian output (sens)

Random number generator used to produce the noise. Seed is based on the time.

Do the MEM instrument spectral radiance convolution of spec (outputs stokes) and all three components of sens1 (output sens_bX)

Read input file:

Radiance simulator with current/dB (spec)

At this point we have stokes0, stokes, and the sensitivity for the dB components. Spectra convolution.

Estimate the “predict” magnetic field (without noise – perfect measurement)

Save the “model” magnetic field (input of the radiance simulator – dB and B from IGRF)

Estimate the “measured” magnetic field (with noise – “real” measurement)

Done exclusively in IDL

SSWRF II

03 - 07 October 2022

Rafael Mesquita (rafael.mesquita@jhuapl.edu)

10 of 21

EZIE Science Questions and Closure Assessment

10

Two-loop model

Sergeev et al., 2014

Large-Scale

Sergeev et al., 2014

Substorm current wedge

McPherron et al., 1974

Double-wedge model

Gjerloev and Hoffman, 2014

Wedgelets

Liu et al., 2018

SSWRF II

03 - 07 October 2022

Rafael Mesquita (rafael.mesquita@jhuapl.edu)

11 of 21

OSSE Results

11

Different examples of OSSES:

“Laminar” current configuration

SSWRF II

03 - 07 October 2022

Rafael Mesquita (rafael.mesquita@jhuapl.edu)

12 of 21

OSSE Results

12

Different examples of OSSES:

Disconnected post- and pre-midnight current wedges

SSWRF II

03 - 07 October 2022

Rafael Mesquita (rafael.mesquita@jhuapl.edu)

13 of 21

OSSE Results

13

Different examples of OSSES:

Non-recurrent/complex current configuration

SSWRF II

03 - 07 October 2022

Rafael Mesquita (rafael.mesquita@jhuapl.edu)

14 of 21

OSSE Results

14

Different examples of OSSES:

Wedgelet

How to go from the line plots back to MHD maps?

SSWRF II

03 - 07 October 2022

Rafael Mesquita (rafael.mesquita@jhuapl.edu)

15 of 21

OSSE Results

15

-300 -200 -100 0 100 200 300 (nT)

Global MHD (GAMERA)

Spherical Elementary Current System Inversion with OSSE

SSWRF II

03 - 07 October 2022

Rafael Mesquita (rafael.mesquita@jhuapl.edu)

16 of 21

Summary

OSSE is a great tool simulate satellite measurement, when used with state of the art atmospheric models to optimize mission design and operations, establish system requirements, test and ready retrieval algorithms, demonstrate measurement capabilities, ensure science closure, etc.
The EZIE OSSE shows, using GAMERA MHD model, that EZIE will be able to close the science objectives.

16

SSWRF II

03 - 07 October 2022

Rafael Mesquita (rafael.mesquita@jhuapl.edu)

17 of 21

References

17

Forward model explanation

Current calculation from EZIE

Measurement technique and sensor design

SSWRF II

03 - 07 October 2022

Rafael Mesquita (rafael.mesquita@jhuapl.edu)

18 of 21

Rafael Mesquita (rafael.mesquita@jhuapl.edu)

19 of 21

OSSE – Basic Info

19

EZIE measurements:

O2 118 GHz emission:

Integrated sub-limb geometry along the line of sight. Peaks at ~80 km with 10 km FWHM.

Magnetic field (B):

Magnitude of B_north, B_east, B_down (𝛿B~75nT) from separation between different polarizations.

Line of sight neutral wind (U):

Doppler shift: ~2 m/s 1-σ precision in 3 s.

Line of sight temperature (T):

Spectral radiances: ~1 K 1-σ precision in 3 s at 40 km.

How do we know this?

-4

-2

0

2

1

Frequency shift (MHz)

14

12

10

8

6

4

2

0

Contribution (K/km)

140

120

100

80

60

40

20

Altitude (km)

Contribution Function (Beam 0, Stokes: V)

SSWRF II

03 - 07 October 2022

Rafael Mesquita (rafael.mesquita@jhuapl.edu)

20 of 21

Observing System Simulation Experiment (OSSE)

20

EZIE’s OSSEs use predicted orbits with optimally designed MEM instrument (expected FOVs, spectral performance, and noises) and EZIE retrieval algorithms

OSSE Step 1: Global MHD Simulations

OSSE Step 2: Observation Event Simulations

OSSE Step 3: Observed Radiances Simulations

OSSE Step 4: Vector B-Fields Retrievals

OSSE Step 5: Current Retrievals

EZIE’s OSSEs assess the retrieved B and current fields to demonstrate science closure.

SSWRF II

03 - 07 October 2022

Rafael Mesquita (rafael.mesquita@jhuapl.edu)

21 of 21

OSSE – Level I Diagram

21

(I) RADIANCE SIMULATOR:

Inputs:
Magnetic perturbations (dBs): Global MHD GAMERA model (high latitude) and WACCM-X/TIEGCM (low latitude)
Background magnetic field (Bs): IGRF/12 and 13 (depending on the OSSE version).
Atmospheric parameters (Temperature, pressure, winds, and density of O, O2, N2…): MSIS (considers winds are equal zero) and WACCM-X (full parameters)
Comment: Done with and without dBs (one run with Bs + dBs and one with only Bs).
Output: Radiance spectra (temperature brightness in Kelvin vs frequency of spectropolarimeter centered at 118 GHZ per point in the orbit per beam)

(II) JACOBIAN CALCULATOR (Sensitivity):

Input:
Currently, the radiance simulator output without the dBs.
Comment: In the future, must be done for the remainder of EZIE measurements: winds, temperatures,…
Output: Sensitivity (Nanotesla per Kelvin vs frequency of spectropolarimeter centered at 118 GHz per point in the orbit per beam).

(III) Retrieval of measurements (Precision)

Input:
Radiance simulator with dBs
Radiance simulator without dBs
Sensitivity (Jacobian)
Comment: Last stop of the OSSE where the noise and error is added to the measurements.
Output: Vector dBs in nanoteslas per point in the orbit per beam.

The steps below are done separately but in order

SSWRF II

03 - 07 October 2022

Rafael Mesquita (rafael.mesquita@jhuapl.edu)