Preliminary Results on the Sensitivity to Radiations of the Back-Up DORIS/Jason Oscillator

Preliminary results on the sensitivity to radiations of the back-up DORIS/Jason oscillator

Pascal Willis Institut Geographique National, Direction Technique, Saint-Mande, France Jet Propulsion Laboratory, California Institute of Technology, Pasadena, USA Context: [email protected] Conclusions: A problem has been detected on the Goals and method: Unfortunately the new DORIS/Jason receiver is also sensitive to JASON/DORIS oscillator when the The goal of this study is to verify if the new DORIS oscillator is also radiations over the SAA. This does not affect the current Precise Orbit satellite crosses the South Atlantic sensitive to radiations over the SAA. We have analyzed time series of Determination (POD) results but it totally forbids any use for geodetic Anomaly (SAA) region. DORIS weekly stations coordinates to look for erroneous velocities applications. Present results (computed using only 2 months or data) On June 29, 2004, the back-up created by the SAA effect and try to compare the velocity before and after show that the amplitude of the effect has an opposite sign but is smaller DORIS receiver (Jason-2) has been June 29, 2004. by a factor of two. turned on

References: J.-M. Lemoine,R. Biancale, A model of DORIS frequency correction for Jason in relation to the SAA, First step Method: presented at the COSPAR meeting, Paris, France, July 2004. We compare for each week and for each station the P. Willis, B. Haines, Y. Bar-sever, et al., TOPEX/Jason combined GPS/DORIS orbit determination in the Analyzing time series of stations coordinates position obtained using the weekly DORIS data with a tandem phase, Adv. Space Res., 31(8), 1941-1946, 2003. on a station by station basis recent cumulative DORIS solution IGN04D02 P. Willis, B. Haines, J.-P. Berthias, et al., Behavior of the DORIS/Jason oscillator over the South Atlantic (position + velocity). Anomaly, CR Geosci, 336(9), 839-846, 2004. Stations inside and around the SAA where not used to transform the weekly solutions into ITRF200. Jason-1 Jason-1 Jason-1 Jason-2 Jason-2 Jason-2 GREENBELT / LATITUDE GREENBELT / LONGITUDE GREENBELT / VERTICAL 150 150 150 Greenbelt (CM) (USA/MD) is far from 100 (CM) 100 (CM) 100 the SAA. Results are not affected 50 50 50

by the SAA effects. RESIDUAL RESIDUAL RESIDUAL Residuals show a 2-3 0 0 0 cm coherence from week to week. -50 -50 -50 POSITION The velocity is totally POSITION POSITION explained by global -100 -100 -100 velocity model derived

using all DORIS WEEKLY -150 WEEKLY -150 WEEKLY -150 satellites for the 1993- 2002 2002.5 2003 2003.5 2004 2004.5 2005 2002 2002.5 2003 2003.5 2004 2004.5 2005 2002 2002.5 2003 2003.5 2004 2004.5 2005 2004.5 period. YEAR YEAR YEAR NB: to transform stations coordinates into ITRF2000 (using a standard 7- The last Jason-1 results (around June 2004) are more noisy because more and more parameter transformation), stations inside the SAA or close by were not DORIS data are rejected during the pre-processing phase at CNES used to estimate the transformation. However, their coordinates were estimated using the derived transformation. Kourou (French Jason-1 Jason-1 Jason-1 Guyana) is inside the Jason-2 Jason-2 Jason-2 SAA region. KOUROU / LATITUDE KOUROU / LONGITUDE KOUROU / VERTICAL Results are affected by 150 150 150 the SAA. (CM) Errors are mostly in (CM) (CM) 100 100 100 latitude as predicted by a simple model (Willis, 50 50 50 2004) RESIDUAL The second DORIS RESIDUAL RESIDUAL 0 receiver (black dots) is 0 0 also affected by the same effect, but with an -50 -50 -50 POSITION opposite sign and a POSITION POSITION smaller amplitude -100 -100 -100 (latitude and longitude). WEEKLY -150 WEEKLY -150 WEEKLY -150 2002 2002.5 2003 2003.5 2004 2004.5 2005 2002 2002.5 2003 2003.5 2004 2004.5 2005 2002 2002.5 2003 2003.5 2004 2004.5 2005 YEAR YEAR YEAR

It must be noted that the effect on the height was up to 1 m even if the DORIS station was not inside the SAA.

Jason-1 Jason-1 Jason-1 Jason-2 Jason-2 Jason-2 LIBREVILLE / VERTICAL Libreville (Gabon) LIBREVILLE / LATITUDE LIBREVILLE / LONGITUDE 150 is outside the SAA but 150 150

in its close periphery. (CM) Geodetic results are also (CM) 100 (CM) 100 100 affected by the SAA effect (mostly the 50 50 50

altitude). RESIDUAL RESIDUAL RESIDUAL The second DORIS 0 0 0 receiver (black dots) is also affected by the -50 -50 -50 POSITION same effect, but with an POSITION POSITION opposite sign and a -100 -100 -100 smaller amplitude (all 3 WEEKLY

components). WEEKLY WEEKLY -150 -150 -150 2002 2002.5 2003 2003.5 2004 2004.5 2005 2002 2002.5 2003 2003.5 2004 2004.5 2005 2002 2002.5 2003 2003.5 2004 2004.5 2005 YEAR YEAR YEAR Second step Analyzing apparent residual velocity per station and per DORIS receiver Analyzing multi-satellite solutions ALL SATELLITES (EXCEPT JASON) ALL SATELLITES (EXCEPT JASON) ALL SATELLITES (EXCEPT JASON) For each station, we have divided the DORIS ESTIMATED DRIFT IN LATITUDE ESTIMATED DRIFT IN LONGITUDE ESTIMATED DRIFT IN VERTICAL times series in two consecutive data sets (before 2000 2000 2000 and after June 29, 2004). 1500 1500 1500 For each data set and each station, we have estimated an apparent drift in all components 1000 1000 1000 (latitude,longitude,altitude) as well as its estimated precision. The plate tectonic effect has 500 500 500 previously been removed. PERIOD PERIOD 0 PERIOD 0 0 Each point of the plot correspond to a unique station -500 -500 -500 X = estimated drift before June 29, 2004 JASON-2 JASON-2 -1000 JASON-2 -1000 -1000 Y = estimated drift after June 29, 2004 Units are in mm/year -1500 -1500 -1500 No erroneous velocity can be detected. Sigma in -2000 -2000 -2000 -2000 -1500 -1000 -500 0 500 1000 1500 2000 -2000 -1500 -1000 -500 0 500 1000 1500 2000 -2000 -1500 -1000 -500 0 500 1000 1500 2000 Y are very large as the secular drift is computed using only 2 months of data. JASON-1 PERIOD JASON-1 PERIOD JASON-1 PERIOD

Analyzing Jason-only solutions

JASON - ONLY JASON - ONLY JASON - ONLY The same analysis done with Jason-only results shows some significant outliers. These ESTIMATED DRIFT IN LATITUDE ESTIMATED DRIFT IN LONGITUDE ESTIMATED DRIFT IN VERTICAL are abnormal velocity detected during the fist 2000 2000 2000 2 years of the DORIS/ Jason-1 that can 1500 1500 1500 directly linked to the SAA effect on the DORIS clock. 1000 1000 1000 By comparing the mean slope Y/X we could 500 500 500 get an idea of the difference of sensitivity to the radiation between the two DORIS 0 0 0 instruments. However, the first 2 months of operation do JASON-2 -500 JASON-2 -500 JASON-2 -500 not seem to be sufficient to derive any -1000 -1000 -1000 reliable estimate of this value. One solution would be to impose that the -1500 -1500 -1500 effect should be 0 in December 2001 (launch -2000 -2000 -2000 of the satellite) but plots above also show that -2000 -1500 -1000 -500 0 500 1000 1500 2000 -2000 -1500 -1000 -500 0 500 1000 1500 2000 -2000 -1500 -1000 -500 0 500 1000 1500 2000 the error does not have a truly linear JASON-1 JASON-1 JASON-1 behavior.