Introduction
With a broad energy coverage from the radiation belt electron source population to hazardous relativistic electrons as well as available pitch angle information for each energy channel, the SST data from the five-probe THEMIS mission has enormous potential for studying radiation belt electron dynamics. While various calibration efforts have continued to improve the absolute accuracy of the SST electron fluxes, in this study we follow the Liouville's theorem to perform empirical cross-satellite calibration of THEMIS SST data based upon electron phase space density (PSD) conjunctions with high-quality LANL SOPA measurements.
|
|
| Figure 1.Hourly averaged electron PSD data inferred from (a) THEMIS-d SST and (b) LANL-01A SOPA data from July 1, 2007 to December 31, 2007. (c) PSD conjunctions within 5.9 < L* < 6.1 for the satellite pair and (d) distribution of PSD conjunctions with normalized difference before calibration. The results after cross-satellite calibration are shown in (e) and (f). μ = 37 MeV/G and K = 0.05 G0.5 RE. Electron PSD is in units of (c/MeV/cm)3 and the dashed diagonal of slope = 1 represents perfect cross-satellite calibration. |
|
Click each image to enlarge.
|
Results
Figure 1 presents an example of inferred electron PSD versus L* (here L* is the Roederer L-shell, equivalently, the third adiabatic invariant) distribution and the performance of cross-instrument calibration for a specific SST energy channel during a half year period from July 1, 2007 to December 31, 2007. The first adiabatic invariant μ is 37 MeV/G and the second adiabatic invariant K is 0.05G0.5 RE, corresponding to the trapped electron population at 66 keV with high pitch angles of ~ 61° for L = 6 in a dipole model. Figures 1(a) and 1(b) show the original hourly averaged electron PSD computed using the T96 model and the measurements from THEMIS-d and LANL-01A, respectively. The PSD conjunctions within 5.9 < L* < 6.1 are shown in Figure 1(c). A histogram of normalized difference (ND) is shown in Figure 1(d), revealing a distribution peaked between -70% and -30% that suggests an underestimate of THEMIS-d SST fluxes compared to the LANL-0A SOPA data. After multiplying the THEMIS-d electron PSD by a factor of 1.62, the scatter of PSD conjunctions and the histogram of the occurrence of PSD conjunctions are presented in Figures 1(e) and 1(f), respectively. The number of PSD conjunctions exhibits a Gaussian variation around ND = 0, indicating an excellent intercalibration of the THEMIS-d SST 66 keV energy channel with the LANL-01A SOPA data.
|
|
| Figure 2. Cross-satellite calibration of THEMIS (a-e) SST data based on comparisons with LANL-01A data from July 1, 2007 to December 31, 2007. K = 0.05 G0.5 RE and PSD conjunctions are obtained using the T96 model and confined to 5.9 < L* < 6.1 for the color-coded μ-values. PSD conjunctions after intercalibration are shown in (a), (c), (e), (g), and (i) for THEMIS a-e with the correlation coefficients in (b), (d), (f), (h), and (j). PSD conjunctions and correlation coefficients between LANL-01A and GEO1989 are shown in (k) and (l). (m) Cross-satellite calibration factors and (n) normalized root mean square errors as a function of energy for each THEMIS probe. Electron PSD is in units of (c/MeV/cm)3. |
|
Click each image to enlarge.
|
With the selected 11 μ-values (from 22 MeV/G to 3581 MeV/G) corresponding to the 11 THEMIS SST energy channels from 40 keV to 2159 keV for the fixed value of K = 0.05 G0.5 RE at L* = 6, cross-satellite calibrations of THEMIS (a-e) SST data are performed based on electron PSD conjunctions with the LANL-01A SOPA, the results of which are shown in Figure 2. Figures 2(a, c, e, g, i, k) show the scatter of joint PSD points for the indicated six satellite pairs with respect to the color-coded μ-value after intercalibration. For each satellite pair the dashed diagonal line of slope = 1 means a perfect agreement. The corresponding energy dependent cross-correlation coefficients (CC) for each of the five satellite pairs, Figures 2(b, d, f, h, j), also indicate that there is a constantly good correlation between the THEMIS and LANL-01A data, particularly for Ek < 1 MeV. Figure 2m shows the obtained intercalibration factors for each THEMIS probe as a function of energy channel and Figure 2n shows the normalized root mean square errors between the LANL-01A and calibrated THEMIS PSD conjunctions. Compared to the LANL-01A SOPA data, the THEMIS SST measurements underestimate the electron fluxes within a factor of 2 for the 40 - 140 keV energy channels and overestimate the electron fluxes within a factor of 3 for the 204 - 2159 keV energy channels.
|
|
| Figure 3. 10-minute averaged PSDs inferred using the T96 model for (a) K = 0.025 G0.5 RE and μ = 700 MeV/G and for (b) K = 0.025 G0.5 RE and μ = 300 MeV/G for THEMIS a-e, LANL-01A, and GEO1989 during a solar wind pressure pulse event. (c) Dst, (d) IMF Bz, (e) solar wind dynamic pressure, (f) the locations of magnetopause (blue curve) and plasmapause (red curve). |
|
Click each image to enlarge.
|
By applying the cross-satellite calibration to the THEMIS SST measurements, we have studied the radial gradient and temporal variation of electron PSD in the magnetosphere to a pronounced solar wind pressure pulse event. Figure 3 shows a clear correlation between electron PSD dropouts and the solar wind pressure pulse, which can be attributed to the combination of losses to the magnetopause and outward radial diffusion. The gradual buildup of electron PSD after the abrupt pressure enhancement likely results from a combination of local acceleration and inward and outward radial diffusion that refills the main-phase PSD dropout during the outward movement of the magnetopause.
Conclusion
The main conclusions of this study are summarized as follows:
(1) The SST instrument aboard each THEMIS spacecraft responds quite similarly to the ambient electron radiation environment. Comparisons of electron PSDs between the THEMIS SST and LANL-01A SOPA measurements show that there is constantly a good correlation between these two datasets and that the differences in electron fluxes are likely due to systematic errors in the THEMIS SST fluxes.
(2) THEMIS SST data underestimate the electron fluxes within a factor of 2 for the 40 - 140 keV energy channels and overestimate the electron fluxes within a factor of 3 for the 204 - 2159 keV energy channels. After cross-satellite calibration the THEMIS-LANL PSD conjunctions show a good agreement particular for energy channels < 1 MeV.
(3) A case study on a pronounced solar wind pressure pulse event using the intercalibrated THEMIS SST data and the LANL and GEO SOPA data demonstrates a clear correlation between electron PSD dropouts and the pressure pulse, presumably owing to a combination of magnetopause shadowing and outward radial diffusion.
Source
Ni, B., Y. Shprits, M. Hartinger, V. Angelopoulos, X. Gu, and D. Larson (2011), Analysis of radiation belt energetic electron phase space density using THEMIS SST measurements: Cross-satellite calibration and a case study, J. Geophys. Res., 116, A03208, doi:10.1029/2010JA016104.
Biographical Note
Binbin Ni is an Assistant Researcher in the Department of Atmospheric and Oceanic Sciences at UCLA. His current research interests focus on resonant wave-particle interactions in the magnetosphere and their quantitative contributions to radiation belt electron dynamics and diffuse auroral precipitation.
Please send comments/suggestions to
Emmanuel Masongsong / emasongsong@igpp.ucla.edu
