COMMENTS.TXT is a newsletter file that will be updated periodically with information concerning the Meteor-3/TOMS, Nimbus-7/TOMS, and future TOMS data sets. The information in this file is informal in nature and may be revised at any time. What responsibility there is for the contents of this file will be borne by Dr. Jay Herman Code 916 Goddard Space Flight Center Greenbelt, MD 20771 herman@tparty.gsfc.nasa.gov ------------------------------------------------------------------------------- 6/4/96 Version 7 of the Meteor-3/TOMS data set is now available. 2/27/95 Since Meteor-3/TOMS is no longer operating, there will be no operational TOMS instruments until the launch of Earth Probe TOMS in May. Until then, the only available daily, full-coverage global ozone data is from NOAA's TOVS instrument. Note: Since TOVS data is produced by NOAA, they are responsible for its quality. INFORMATION ON OBTAINING THE TOVS TOTAL OZONE DATA GRIDDED BY CLIMATE ANALYSIS CENTER OVER INTERNET. The daily, global analyses of total ozone derived from the NESDIS TOVS instrument (currently NOAA-11, soon to be NOAA-14) are available via Internet at about 12:30 EST for the previous day on the Workstation known as the National Information Center (NIC). The data are at a 1 x 1 degree resolution on GRIB format. The data are being located in two places for accessibility; the Internet address and on MOSAIC. Directions for obtaining the data are given below: INTERNET- ftp nic.fb4.noaa.gov user: anonymous password: your email address cd pub/nws/nmc/cac/htdocs/products/stratosphere/tovsto/ ls lists the file available file name contains date of analysis e.g. 950213.grb use binary mode to transfer get file quit The file size is 179,274 bytes. Documents and codes for Grib available on the NIC at: pub/nws/nmc/docs/README pub/nws/nmc/docs/gribguide pub/nws/nmc/docs/gribed1 pub/nws/nmc/codes MOSAIC- address: http://nic.fb4.noaa.gov/products/stratosphere/tovsto In case of problems contact Jim Miller at miller@upair.wwb.noaa.gov --------------------------------------------------------------------- TOMS COMMENTS 2/27/95 METEOR-3/TOMS is still not working. A final restart attempt will be made in mid-April. After that, we expect the new TOMS to become operational soon after its May 1995 launch. This is a 1 month slip from the guess made in January. 1/20/95 So far efforts to restart Meteor-3 have been unsuccessful. The cooperation from the Russians has been excellent in performing the few possible operations that might help restart TOMS. We anticipate that the final attempts will be on April 15, 1995 when Meteor-3/TOMS is at its maximum temperature. If the instrument does not start at this time, we expect to shut it down. Hopefully, the launch of Earth-Probe Toms will occur successfully at this time. Current plans call for a launch between April 10 and April 20. After launch, the first TOMS data will start in about 30 days. 1/9/95 Restart efforts for M3/TOMS will begn on January 12, 1995. Initially this will consist of our Russian partners heating their spacecraft to 35C, and therefore TOMS to about 25C to 30C. This is in an effort to free the bearing in the TOMS chopper motor. At the present time the rest of TOMS, the spectrometer and electronics, are functioning. We are receiving some data from one of the six wavelength channels. The data are not of sufficient quality to produce scientifically interesting results (ozone or radiances). 1/3/95 METEOR-3/TOMS HAS CEASED OPERATION ON DECEMBER 27, 1994 DUE TO FAILURE OF THE CHOPPER MOTOR. THAT IS, THE CHOPPER MOTOR CURRENT IS ZERO EXCEPT FOR INFREQUENT SPIKES TO 70 MA. AT THE PRESENT TIME THERE ARE NO FURTHER TOMS DATA AVAILABLE. IF EARTH-PROBE TOMS IS LAUNCHED, THE EARLIEST DATA WILL BE DURING JUNE 1995. IF EARTH-PROBE TOMS IS NOT LAUNCHED, THEN THE NEXT TOMS LAUNCH WILL BE ABOARD A JAPANESE SATELLITE, ADEOS, SOMETIME DURING 1996. THE EARLIEST DATA FROM ADEOS/TOMS WILL BE DURING APRIL 1996 IF THERE ARE NO DELAYS. EFFORTS ARE BEING MADE TO REVIVE METEOR-3/TOMS. 12/29/94 Because of the chopper motor, Meteor-3/TOMS has not produced any good data for the 28th and, possibly, the 29th. Unfortunately the problem shows no sign of improving. 12/28/94 Meteor-3 is having a lot of trouble. The chopper motor problem is the reason we have very large data gaps and days with no data during the past few days. Unfortunately the problem shows no sign of improving. 12/27/94 Meteor-3 continues to obtain increased coverage of the northern hemisphere as it drifts closer to a noontime equator crossing. The problem with the chopper motor is causing large data gaps and, because of it, we have no data for December 24, 1994. This problem seems to be steadily getting worse. 12/21/94 We are losing data, once again, because of the chopper motor sync problem. Meteor-3 is presently experiencing high temperatures. High temperatures, like low temperatures, may put a strain on the instrument, and we hope the sync problem goes away as the satellite returns to more moderate temperatures. 12/13/94 Along with the chopper sync problem, we have had problems with the transmitter on the satellite. As a result, we do not have data for 2-Dec to 5-Dec. Meteor 3 is now operating on one of its other transmitters, but we are still experiencing some data transmission problems. This means that we may have to wait for the second data dump (we get two each day) before we process the data, and our processing may therefore not be as timely as we would like. We hope, using the other transmitter, not to lose too much data. 12/1/94 METEOR-3 is experiencing problems on almost every day that cause data loss. It appears as if the chopper motor synchronization problem is becoming worse than before. The satellite orbit is also nearing the terminator (every 106 days) where the data quality is reduced because of the high solar zenith angles (see the message of 5/5/94). 11/29/94 We've managed to partially recover data for the 23rd and 24th of November. At least for now, however, the quality of data for these two days is not as good as data for other days. We are still investigating what happenned. 11/28/94 There was a problem with data downloaded from Meteor-3/TOMS for the 23nd and the 24th of November. In fact, the data we received was impossible to process and we therefore only have partial coverage for the 23rd and no coverage for the 24th. This problem also affected subsequent data processing for the 25th through the 27th. We are attempting to bypass the two problem days in order to process the 25th, 26th, 27th, and later days. When we obtain more information as to what happened on the 23rd and 24th, we will post it here. 11/14/94 November 11 was the worst day, in terms of obtaining good data, we've had since June, 1993. We are still evaluating data rejected because of the chopper motor sync problem to see if it can be used, but the early analysis of it does not look promising. 11/04/94 For the past month Meteor-3/TOMS has been experiencing increased data rejection due to the chopper motor sync problem. We are presently trying to evaluate the data that has been rejected to see if we can somehow salvage it. We also hope that, as the spacecraft temperature rises, the problem will be alleviated. 09/02/94 Meteor-3/TOMS has had some transmitter problems that appear to be corrected now. We had no data loss because of these problems. This year at the approach to the ozone hole season in Antarctica the coverage will be late. This is because the spacecraft's orbit is along the day-night terminator and the southern hemisphere is still pointed away from the sun. The combination puts the satellite into the night side in the southern hemisphere where we can not obtain data. By mid-September Meteor-3/TOMS should get full global coverage. We have examined the available data for 1994 and compared it to data 1993 and before. The data are in good agreement with previous years within the expected variance. The data have also been compared with US Dobson stations and are in good agreement except for Instrument 83 located at Mauna Loa, Hawaii. M3/TOMS agrees with Instrument 83 when located at Boulder, Colorado. Meteor-3/TOMS data are about to undergo a reprocessing that includes a new examination and review of the in-flight calibration. This should improve the data quality. The M3 reprocessing will follow the reprocessing of Nimbus-7/TOMS with the new Version 7 algorithm. I will supply some details of this algorithm during the next month or so. Some new information about Meteor-3/TOMS is contained in a publication about to appear in J. Geophys. Res. Vol. 99 1994 by Herman et al., "Meteor-3/TOMS Observations of the 1993 Ozone Hole". We are expecting the launch of a new TOMS instrument sometime between November and February. Earth-Probe TOMS will be in a sun synchronous orbit similar to Nimbus-7/TOMS. It will have improved in-flight calibration capabilities, an additional short wavelength channel, and 1 less long wavelength channel (380 nm). More details will be provided after launch. 07/20/94 Our near real-time production is operating again, and we should be caught up shortly. 07/19/94 The disk that our production system resides on died over the weekend and, as a result, we have not produced any data since Saturday. We are working towards resuming production sometime today or tomorrow. 05/30/94 NIMBUS-7/TOMS data will be reprocessed this summer with a new algorithm that will remove some problems due to initial ground and subsequent in-flight calibration errors. The new algorithm will be applied to the Meteor-3/TOMS data and the future Earth-Probe/TOMS data (launch mid-summer 1994). The EP/TOMS will be a new sun-synchronous TOMS. 05/25/94 We have once again changed the directory structure for Meteor-3/TOMS; the directories yYY (where YY is the year) are now called ozYY (08/23/91 - present); the directory rf94 (1/1/94 - present) contains files of the reflectivity calculated from Meteor-3/TOMS (the reflectivity is an average of the 360 and 380 nm channels). CAUTION: A number of assumptions are made in the calculation of this reflectivity (that the reflecting surface is uniformly Lambertian, that there are no aerosols,etc). When these assumptions are not valid (eg in the presence of aerosols) or begin to break down (eg when Meteor nears the terminator) the reflectivity calculation is strongly affected and can contribute to errors in ozone determination. 05/05/94 The approximate dates of maximum data loss in the northern hemisphere are: NORTH SOUTH The period of substantial data loss is about 1.5 weeks on either side of the 9/27/91 maximum day. From the dates for both 1/13/92 hemispheres, the maxima are separated 4/25/92 by roughly 3.5 months corresponding to 8/9/92 the 106 day half-precessional period of 11/29/92 M3/TOMS. Maxima that occur during 3/13/93 solstice are more complete than those 7/1/93 occuring near equinox. This is because 10/13/93 the satellite looks sideways into 1/24/94; daylight as it travels along the 5/15/94 terminator. The apparent irregularity 9/1/94 is caused by the relationship between 12/15/94 the orbit's precession, the day-night 3/30/95 terminator, and the seasonal tilt of 7/15/95 the earth's spin axis relative to the 11/2/95 plane of the ecliptic. 2/10/96 We never lose ozone data during the development of the Antarctic ozone hole. This is because it occurs during equinox when TOMS can look sideways into the daylight and completely cover the polar region every day. 04/28/94 We are back to full data coverage. We have been working on making GIF images of the data. These images may soon appear in a directory under images called gif. We are thinking about only carrying one week's worth of PostScript images (which are quite large) and converting the rest into GIF format (which are much more compact and easy to manage). 03/16/94 We are experiencing some data loss due to the chopper motor problem discussed previously. This problem is somewhat correlated with spacecraft temperature, and Meteor-3 is close to reaching the coldest temperatures it experiences in its orbital precession. As the spacecraft begins to warm, we hope the problem will alleviate. 02/07/94 We have solved the problems we were having due to the upgrade of IDL. We will now be providing images as soon as the data is produced. 02/04/94 Our near real time processing system has broken down due to an upgrade in the software we use (IDL) to produce the PostScript images. Our software crashes using the new version of IDL (which is incompatible in certain places with the old version) and our whole real time system therefore comes to a stop. It will take some time to work through the incompatibility problems between the old and new versions of IDL. In the meantime, we will change the system to provide the data as soon as possible and will attempt to provide the images using a machine that hasn't been upgraded yet. 01/27/94 We have instituted a number of changes in the structure and content of this ftp account. First, the naming convention of the Meteor-3/TOMS data sets has been changed. From now on the files will be named gaYYMMDD.m3t where YY stands for year, MM stands for month, and DD stands for day. For example, ga940125.m3t will be the data for January 25, 1994. The g in the name stands for gridded data and the a signifies the fact that these are ASCII files. The change was done to bring the naming convention more in line with that of Nimbus-7/TOMS. Also, we have separated different years into different subdirectories under meteor3. Second, we have placed data back to the launch of Meteor-3/TOMS into these directories. The data before July, 1992 was processed before our real-time system was developed and therefore has a slightly different header that declares the data to be Version 0.9. All data contained in this account is good quality data. Finally, to save space we are now producing only one PostScript image file containing all three views. These image files will be placed in the image subdirectory under the meteor3 directory. The naming convention for these files will be miYYMMDD.ps where m stands for Meteor-3 and i stands for image. 01/26/94 We have now replaced data from 1-Jun-93 (julian day 152) to the present with the reprocessed data (data before 1-Jun-93 was not affected by the reprocessing). We also will soon make available data (in CDTOMS format) taken since the launch of Meteor-3/TOMS. In the near future, we plan to do a little housekeeping and rearranging. First of all, the naming convention of the data will be changed to bring it more in line with that of Nimbus-7/TOMS. The exact naming convention hasn't been decided yet. The decision should be made today, with the change implemented by the end of the week. The second change we are planning affects the posting of PostScript images. These images are large and have taken up a lot of the disk space available for our anonymous ftp account. In an effort to save space while at the same time making images available, we will be consolidating the three separate image views into one file. These images will be placed into a directory under meteor3 called image. The directories world, npolar, and spolar will cease to exist. 01/14/94 M3/TOMS continues to operate normally. The chopper motor continues to be slightly out of synchronization with the photon counting electronics, but the error remains small. Within the limitation imposed by the Meteor orbit, we continue to obtain full global coverage of ozone amounts. One key result is that the low values see at most latitudes during 1992 and 3/4's of 1993 have returned to historically normal values. The reprocessing of Meteor-3/TOMS ozone data is complete. The reprocessing covered the period June 1993 to the present. The main change was in incorporating an in-flight determination of changes in the instrument sensitivity (ACF or Albedo Calibration Factor) after the June 1993 loss of data. We are reviewing the results of the reprocessing, and will post the data in this account on January 24 (Monday) if there are no glaring errors. One of the big changes between the old and new data is the value of the minimum observed ozone amount during the Antarctic ozone hole season. We now measure about 88 DU on October 8, 1993. A detailed publication using the new data is nearly ready for submission to J. Geophysical Research. 11/29/93 The first reprocessing of the Meteor-3/TOMS data is scheduled to begin today and be completed in 2 to 3 weeks. The data will be archived as version 1.0 and will be available on this account. I will present a full description of all changes made to the data from the preliminary version. Nimbus-7/TOMS data will also be reprocessed in about 1 year with major changes to the algorithm and in-flight calibration determination. The new results are expected to contain small, but significant changes in ozone values. The most significant changes will arise from the removal of artifacts introduced by errors in the in-flight calibration of Nimbus and Meteor/TOMS. For Nimbus, this will be version 7.0 and for Meteor version 2.0. By the time this reprocessing is finished, the new Earth-Probe/TOMS is expected to be flying for 9 months. 11/8/93 We are preparing to reprocess the Meteor-3/TOMS data with improved in-flight calibration constants. Since launch in August of 1991, the TOMS instrument makes regular measurements of the solar flux using one of its 3 diffuser plates. The plates differ in the amount of exposure to solar uv and space to limit the rate of degradation on 2 of the 3 plates. On the Meteor spacecraft, the solar calibration has been more complicated than originally anticipated. The complications arise from the precessing orbit causing changing viewing angles, and changing temperatures. We now feel that we have a much better estimate of the changes in instrument sensitivity during the past 2 years of flight using the solar calabration measurements. The data will be reprocessed using the new albedo calibration factor (ACF) during the next few weeks. This data will be archived as Meteor-3 Version 1.0. There will still be some known, but unresolved problems at the time of the reprocessing. The known deficiencies will be discussed thoroughly in this space and in a refereed publication accompanying Version 1.0. We expect to be able to correct the major remaining problems using in-flight techniques only, and using external data for validation. 9/30/93 THE REPLACEMENT OF THE C-PAIR DATA IS COMPLETE FROM 8/15/93 TO THE PRESENT. THIS IS A TEMPORARY REPAIR UNTIL THE SPACECRAFT CALIBRATION IS BETTER UNDERSTOOD. DATA PRIOR TO 8/15/93 DID NOT HAVE CALIBRATION PROBLEMS TO THE SAME EXTENT. 9/30/93 The data in the anonymous ftp account will be replaced from August 15 to the present with B-pair ozone. We feel that these data better represent the actual amount of ozone present in the atmosphere. A review of the C-pair solar calibration data shows that the chopper motor out-of-synch (see below) problem is adversly affecting the c-pair more than other pairs. More on this in a few days. Current balloon sond data are showing extremely low ozone amounts at the south pole in reasonable agreement with TOMS B-pair data. 9/29/93 We have been performing radiative transfer studies to validate the ozone measurements obtained at very high solar zenith angles typical of the Meteor-3 orbit when it nears the terminator. This is the condition this year for the Antarctic ozone hole development. the result of these studies is that the C-pair ordinarily used in the polar region is incorrect by substantial amounts. The c-pair results are currently showing low values of about 95 DU. The same radiative transfer studies also show that the B-pair is quite accurate for high solar zenith angles. We expect to reprocess the ozone data from August 1 to the present in the next few days, and from January 1 to July 31, 1993 shortly thereafter. The reprocessed data will use only the A-pair and B-pair to calculate ozone values, and will also incorporate the recently derived in-flight instrument calibration constants. The later produces a small 1 to 2% correction. The C-pair problem was not present in the Nimbus-7/TOMS data because of its superior near sun-synchronous orbit. Based on the above mentioned studies, the very low values of 95 DU are really about 110 DU. Earlier data was more accurate before the satellite neared the terminator. 9/24/93 The current low south polar ozone value seen by Meteor-3/TOMS is 106 DU on 9/22/93. 9/23/93 Meteor-3/TOMS is reporting low ozone values in the southern hemisphere and the development of the usual Antarctic ozone hole. Minimum values are now below 110 DU within the south polar vortex region. Low values measured outside of the vortex region are consistent with values measured by the Dobson station in New Zealand. Recent comparisons of Meteor-3/TOMS data with the standard Dobson ground station at Mauna Loa show good agreement (i.e., the same +4% to +4.5% TOMS - DOBSON offset persists). The chopper motor out-of-synch condition, discussed here previously, appears not to have affected the mean value of a set of measurements. However, the standard deviation of the set has increased compared to the period before the onset of the out-of-synch condition. There is an offset in ozone values in the overlap regions between the 3 different wavelength pairs used for different solar zenith angle conditions. A-pair (0 - 60 degrees) B-pair (55 - 79 degrees), and C-pair (73 - 90 degrees). The pairs are averaged with linear weighting in the overlap regions. The offset has been changing with time. Relative to A-pair, the B-pair is currently about 3% high, and the C-pair about 5.5% low. If these error estimates prove to be accurate, then the 110 DU measured over the polar regions should be corrected to 116 DU. We are working to improve these error estimates. Comparisons of A-pair with ground based Dobson indicates that TOMS C-pair may be another 5 DU high in the south polar region. The current best estimate of the ozone amount seen by Meteor-3/TOMS is 111 DU. Since there are offsetting errors,the absolute ozone error appears to be approximately 1 DU too low. Given the uncertainty in the measurement, the offset is negligible. It should also be noted that the Meteor-3 orbit is approaching the terminator. Previous comparisons with Nimbus-7/TOMS indicate that there can be an uncertainty of +-3.5% under these conditions. If the known errors are combined, Meteor-3 is measuring 110 +- 4.5 DU. This isn't the final word on Meteor-3/TOMS errors or offset calculations. What is described above is the correction needed to offset the known changes in the instrument's electronics based on in-flight solar calibrations. Unlike the nearly sun-synchronous Nimbus-7/TOMS, Meteor-3/TOMS is very complicated because of its orbit. The entire 1992 data set from Nimbus-7/TOMS has been added to /pub/nimbus7. The corresponding Meteor data will be added soon. 9/9/93 The Meteor-3/TOMS is working normally. Current ozone data is showing a steady drop in the Antarctic region. The day-night terminator is at about -82.5 degrees and ozone amounts are between 170 and 180 DU. The new calibration constants still have not been applied to the data. This means that the ozone values in the polar region are probably higher than we are reporting, and lower at latitudes towards the equator from -60 degrees. Errors are probably on the order of -10 DU in the polar region and +5 DU at midlatitudes. The new calibration constants will be reviewed during the next two weeks - hopefully prior to the minimum in the Antarctic region on about October 1 to October 10. There are some IDL mapping routines in the meteor directory (e.g., polar.pro, minmax.pro, etc.) Their basic function is documented in a single line at the top. If you own IDL, please feel free to use them in your work. We cannot provide any support for the code. 9/2/93 We are now supporting the Antarctic real-time research. Accordingly, postscript files will be loaded into anonymous FTP as part of routine processing. This means that the files will be approximately 3 hours old from time of reception from the spacecraft at Wallops Island, Virginia. Since data reception can occur any time of the day or night, the process has been completely automated. The postscript files and data files are posted in the ftp account with no review. A review both data and image files will occur the next working day. Defective files will be replaced. Almost all of the time this will result in the most recent data being made available in the shortest possible time. 9/1/93 First results from the error analysis indicate that the increased noise after the onset of chopper non-synch has been small. A-pair uncertainty has increased from 0.6% to 0.7%, and B-pair uncertainty from 1.2% to 1.6%. A-pair is used for solar zenith angles between 0 and 55 degrees. B-pair is used at 65 degrees and above. These uncertainties are on top of in-flight solar calibration errors. These in-flight calibration errors have not been corrected as yet. We are working on it. So far it appears as if the A-pair is about 1.5% too low in August and that the B-pair is 1.5% too high in August relative to 1992 values. 8/31/93 We have just completed a first study of error estimates for Meteor-3/TOMS data. I am putting together a summary of the results and will have it posted here in a few days. I'll probably put it into a separate file (e.g. ERRORS.TXT). At that time, we will relax the warning about publication to something more reasonable. It will still be "real-time" data, but we will give some idea its accuracy. Polar postscript maps are now available. We are going to try to produce them in "real-time" instead of sometimes 2 days behind. This means they will be initially uploaded without review, and then perhaps revised later if there are problems. The same is being done with the gridded data. We expect to present some of the error analysis and the development of the Antarctic ozone hole at the Fall AGU meeting in San Francisco. We have been given a warning that the Wallops Island ground receiving station may be down for 1 to 2 days due to hurricane Emily. If so, we will lose the real time data, but may be able to recover the missing days from the Russian receiving station. 8/25/93 The anonymous ftp is now permanently on jwocky in a 500 mbyte partition. jwocky is a newer faster machine on which we have much more disk space than before. This will enable us to have much more data on-line for TOMS. We will now start to backfill the files to August 1991 for Meteor-3 and to January 1, 1992 for Nimbus-7/TOMS. We are not planning to backfill the postscript files in the directories world, spolar, npolar. 8/24/93 Meteor-3/TOMS experienced a minor data loss last week, a small portion of 1 orbit on one day. Overall the data was not effected too badly. We have noticed an increase in apparent noise in the data compared to earlier data. During the next few weeks we are going to try to estimate the accuracy of the data in this account so that it may be used for some types of scientific analysis and perhaps publication (with caution). We are now providing postscript color maps of the ozone amounts on a rectangular grid (pub/world), and may provide polar maps (pub/spolar and pub/npolar). With the apparent loss of NOAA-13 and the SBUV-2 instrument on-board, the Meteor-3/TOMS is the only instrument likely to provide complete coverage until the launch of Earth-Probe/TOMS next spring. With luck we will get some overlap between the instruments to ease the transfer of calibration. 8/16/93 Meteor-3 continues to produce reliable data. A complete view of the south polar region should be available in a couple of weeks as the orbit continues to precess. Please note that we are creating a postscript file that contains an ozone map of the entire visible globe. Please switch to the new machine jwocky. 128.183.58.65 Data will soon be removed from hyperion. 128.183.58.20 7/6/93 We have now regained 100% data coverage from Meteor-3/TOMS. Data since June 22 is (or should be) available as of today. If there are no further problems, we will continue the data record. We also intend to place the entire data set from January 1, 1993 to the present on anonymous FTP for a at least a month (perhaps longer if disk space permits). We have done a rough analysis of the M3 data since the June 2 to June 21 dropout. It appears that the continuation of the data from June 22 joins smoothly with data from June 1 and before. A more thorough analysis involving a review of the instrument calibration is in progress. The instrument is basically in the condition it was in around May 15. The chopper motor synchronization error flag is still on. The error is small, and appears to show up as a small increase in random noise. Please note that the scalloping effect in the southern hemisphere is a result of the precessing orbit being near the day-night terminator. This also means that the ozone values near the scalloped edges are of lower accuracy because of the higher solar zenith angles ( greater than 82 degrees ). We currently expect to have coverage of the Antarctic ozone hole starting in August. However, we will again be near the terminator in early October, but with full global coverage. The accuracy should be adequate for observing the development of the ozone hole. 6/21/93 The chopper motor out-of-synch problem has continued on Meteor-3/TOMS. In addition, the instrument started to lose data starting June 2. Since then we have had only sporadic coverage of a partial nature. The coverage rates have ranged from 0 to 75%. The large rates of data loss appear to be correlated with solar flare activity. There is on-board evidence that charged particle activity has affected some of the TOMS circuitry. In addition, there appears to be indications of increased mechanical friction (higher currents) in the chopper motor circuits. The increased friction could be the cause of the original out-of-synch flag problem. The increased currents coincided with a period of exceptionally low spacecraft temperatures (15C) that correspond to an unusual orbital geometry (instrument facing away from the sun near northern summer solstice). At the present time the spacecraft temperature is rising again back towards normal operating levels (22C). There is reasonable hope that the TOMS instrument onboard Meteor-3 will recover. 5/28/93 Meteor-3/TOMS has experienced chopper motor out-of-synch error flags (CM-flags) starting about May 15. However, the quality of the data is still quite good. Starting with May 22 to May 26, the data have the following percentage of CM-flags 60%, 15%, 60%, 85%, and 85%, respectively. Prior to May 22, the percentage of CM-flags was small (1% - 10%). Except for May 22, maps of the M3/TOMS ozone data are nearly complete. On May 22, the worst day, we lost about 10 to 15% of the total data. Of the remaining 85% to 90% of the ozone data, the out-of-synch data differs by no more than 2% from the in-synch data. Most of the differences are less than 1%. The comparison was made by forming zonal averages around 10 degree wide latitude bands separately for the two classes of data (in-synch and out-of-synch). The chopper wheel is used to automatically correct for the background counts seen by the spectrometer. In order to do this, the rotational position of the slits in the chopper wheel must be synchronized precisely with the counting electronics. CM-flags can be triggered at a level where the counting error is below the threshold for affecting data quality. If the synchronization error is sufficiently out of range, then the software used to convert counts into ozone amounts will detect the error and return null values (i.e., missing data). We are evaluating whether the differences in the 2 classes of data could have any significant implications for using M3/TOMS data to extend the Nimbus-7/TOMS ozone data for use in determining ozone trends. We have already determined that the differences are not significant for the daily ozone maps. Errors of this nature occurred for Nimbus-7/TOMS over an extended period of time. There was very little data loss for N7/TOMS, and the quality was not significantly compromised. 5/7/93 FRIDAY NIMBUS-7 IS DEAD OCTOBER 24, 1978 TO MAY 6, 1993 LONG LIVE METEOR AUGUST 15, 1991 A WAKE WILL BE HELD FOR IMMEDIATE RELATIVES AND OTHER INTERESTED PARTIES DONATIONS OF SCIENTIFIC ARTICLES SHOULD BE SENT TO YOUR FAVORITE JOURNAL 5/1/93 We have begun to produce near real-time Nimbus-7 data on the workstation platform. There are two immediate benefits to this: 1) We should now be able to produce a more timely product. 2) The data gap problem has been eliminated as far as our processing is concerned. Please note that, in the interest of providing data as quickly as possible, some gaps will appear due to problems with the data that have been delivered to us. Rather than wait for a resolution to these problems, we have decided to process what we have. For example, the data delivered to us for April 30 are missing the time period from approximately 1200 to 1500 hours GMT, either due to data not being delivered to us or a problem with the data that was delivered to us. These types of problems will be corrected later in reprocessing. 4/23/93 We are still testing some aspects of our improved Nimbus-7 real-time system. It will be about a week (maybe a little more) before the improved data starts to appear in the nimbus7 directory. It will take a few weeks after that before we reprocess Nimbus-7 data from this year and make it available. 4/22/93 I omitted mentioning that Nimbus-7 is again running without solar calibration. Because of the slowly drifting nearly sun- synchronous polar orbit, the equator crossing time has moved to about 10:40 am (from noon 14-years ago). The result is that the diffuser plate used to look at the sun is now partly in shadow. We usually use the sun to monitor the gain of the photomultiplier tube (PMT) and other electronics associated with TOMS. At present the instrument is operating with an extrapolated calibration curve. However, the in-flight extrapolated calibration has been verified by comparison with the fully calibrated Meteor-3 TOMS, the NOAA-11 SBUV instrument, and ground-based Dobson stations. We plan to magetically rotate the Nimbus satellite this summer so that solar calibrations can again be obtained. This will mean losing about 1 orbit's worth of data every two weeks in exchange for the calibration. Two papers have been accepted for publication (JGR) that bear on data quality from Nimbus-7/TOMS during the period of aerosol contamination from the Mt. Pinatubo eruption. These are by Bhartia et al. and by Herman et al. (1993). I'll give the full references when they are published. The short conclusion is that the aerosols affected the ozone data mostly in the equatorial region, -25 to +25 degrees latitude, in that ozone values obtained from scan positions near 1 and 35 differ from those obtained near the center (17). The effect has been explained quantitatively using radiative transfer calculations that include Mie scattering. Zonal averages are not effected at the 1% accuracy level. Work is in progress to obtain the effects at higher latitudes. 4/21/93 Thanks to the efforts of several of our staff, we have improved the real-time processing system for Nimbus-7/TOMS. Most of the data that appeared on the FTP account had missing sections. The missing data was due to a faulty software program used to strip data from the satellite telemetry stream. The problem occurred when we switched computers (IBM mainframe to VAX to SGI workstations) and discovered that the documentation for a 14-year old satellite, with an expected lifetime of 2 years, was hard to find. We found the dusty files, reprogrammed the software, and should be back in business. We plan to reprocess all of the Nimbus-7 data from Jan 1, 1993 to remove the data gaps. If we have the space, we will put the corrected data sets on anonymous FTP for a few weeks in a separate directory. 4/20/93 Once again, we think we have solved the problems with our computer system. 4/17/93 We have run into problems with our processing system once again. We hope to be up by Monday morning. 4/13/93 We have automated our near real-time stuff for Meteor-3/TOMS to the point that we can provide data immediately after we receive it. Due to the nature of the data from Nimbus-7, we have not been able to automate near real-time production for that satellite nearly as easily. We are working on it (as well as solving the data gap problem for Nimbus-7/TOMS). Stay tuned... Repairs have been completed to our main processing computer. Hopefully, this will mean that we can resume our normal schedule for processing real-time Nimbus and Meteor data. 4/8/93 We are back up with near real-time Nimbus production. Our computer system is causing problems at the moment, however, and we might not be as timely as we'd like to be with Nimbus data for a while. We are also having some problems processing days 90 and 91 and are working on the problem. 4/5/93 We are experiencing a delay in production of real-time Nimbus data due to the fact that the computer we are using is having disk problems. We have no estimate as to how long it will take to resolve the situation. We hope to be up again soon. 3/24/93 Just to get started with Nimbus-7, we have included a longer period of time (starting on Feb 1) than 3 weeks. This will be gradually cut back to the promised 3 weeks. Please note, the production data does not have the gaps present in the real-time data. We will make the entire data set of corrected Nimbus-7/TOMS data available when our problems transferring to new computers are solved. This will probably be sometime during June. 03/09/93 NIMBUS-7/TOMS DATA should be available in the directory pub/nimbus7 The first data sets may be a bit rough because of a defective data stripping routine. We originally wrote this software for an IBM mainframe using proprietary routines (a mistake in 1978). Converting this ancient software to process the data on modern machines has proven to be a more difficult task than originally imagined. The effect of our initial software is to leave gaps in the data when it fails to detect the beginning of a record. This problem only applies to the Nimbus-7/TOMS data, and not to the Meteor-3/TOMS. At the present, REAL-TIME Nimbus-7/TOMS data differs from the production in the following ways. 1. Predicted instead of definitive ephemeris is used. 2. Satellite housekeeping and TOMS status information is not available. 3. The satellite attitude information is not used. 4. There is no review or quality control applied to the data. Items 2, 3, and 4 will be improved as time progresses, but will never be up to the level of production quality data. Also, the production data has no gaps that arise from the data stripping routines. 2/26/93 We are preparing a publication on the 1992 ozone data from Nimbus-7/TOMS that contains a comparison of M3/TOMS with N7/TOMS ozone amounts. The comparison shows that the 2 instruments are very close in value (about 1% absolute) and track each other very well. This is encouraging to us from the standpoint of both N7 and M3 independent in-flight calibrations. The data are also compared with THE NOAA SBUV-2 and with the Dobson ground stations. The results show good agreement as has been reported previously in published work (Herman et al.,1991; McPeters et al., 1991). We are planning to release the real-time Nimbus-7/TOMS ozone data soon. See the file NIMBUS7.TOMS for details. 02/09/93 The Meteor-3/TOMS data have recently been compared with the Nimbus-7/TOMS ozone data in several latitude bands. The results show that the 2 data sets agree to within less than 1% when the Meteor-3 equator crossing times are comparable to Nimbus-7. Currently the Nimbus-7 spacecraft has drifted from its original noon launch orbit in 1978 to about 10:45 am in 1993. Comparable observing times for Meteor-3 are when the equator crossing times are between 10 am and 2 pm. Under these conditions the solar zenith angles are similar for the 2 spacecraft. High solar zenith angles affect the reflectivity determination for the surface of the earth and the observing angle relative to any stratospheric aerosol phase function (e.g., the ozone data obtained for several months after the eruption of Mt. Pinatubo in June 1991). Nimbus ozone data is still in the works. We now expect mid February to be the starting date. There are problems with the Nimbus-7/TOMS raw data stream. Meteor-3 is a more modern setup, and was simpler to work with. Nimbus-7/TOMS required more manual intervention in the production processing. Automating this for REAL-TIME processing is taking a bit longer than expected. REAL-TIME will be missing a few features of the production data at first. Most important among these is the lack of spacecraft attitude information. Most features of the production system will be added as time permits. The goal is to make the production and REAL-TIME identical except for the ephemeris (spacecraft location) and the quality control on the data output. We expect to use the REAL-TIME data to monitor the quality of the received data so as to detect problems in time to make corrections. 2/10/93 Up to this point, the ascending node local equator crossing times supplied in the header of the Meteor-3 files have been approximated by subtracting 7 minutes to each day. This has lead to an error of about 15 minutes in the latest data. A new procedure has been developed to include a more accurate calculation of this time. The data currently in this directory (from 1/18/93 to 2/7/93) have been redone to include the more accurate times in the header, and future processing will be done using the more accurate procedure. Remember, an ascending node local equator crossing time in darkness (eg 11pm) corresponds to a descending node local equator crossing time in daylight (eg 11am). 01/14/93 Currently only the Meteor-3/TOMS gridded ozone data are available on anonymous FTP. Within a few weeks (perhaps 2 weeks) the real-time version of Nimbus-7/TOMS will also be available. Comparisons have been made between the real-time M3 data and the production data produced 3 - 4 weeks later. The difference is less than 1% over the entire globe. The difference arises from the use of an updated ephemeris based on orbital state vectors supplied by the Russians (Meteor-3/TOMS only). Because of the precessing nature of the Meteor-3 orbit (precession period = 212 days), the equator crossing time goes from noon (overhead sun, solar zenith angle = 0 deg) to the terminator (solar zenith angle = 90 deg). The best data quality is when the equator crossing time of the ascending node of the orbit is between 10am and 2pm or between 10pm and 2am (which indicates that the equator crossing time of the descending node of the orbit is between 10am and 2pm). Good data quality is obtained over a wider range of equator crossing times, but the quality has not been put on a quantitative basis as yet. Until I put a chart with more accurate times, the equator crossing times can be estimated from straight lines connecting the following pairs of points. Days since launch Equator crossing time on 08/15/91 (hours) 0 11 96 0 96 24 250 7 Some other potentially useful data concerning the Meteor orbit and Meteor-3/TOMS characteristics. altitude 1202 km orbit inclination 82.5 deg orbit period 109 min orbit eccentricity < 2 x 10^-3 orbit precession 212 days launch date 08/15/91 launch time 12:15 Moscow standard time 1st ozone data 08/22/91 # orbits/day 13.21 field of view 3 x 3 degrees wavelengths 312.35, 317.4, 331.13 339.73, 359.0, 380.16 nm bandwidth 1.1 nm cross-track scan angle +/- 51 deg number of views/scan 35 nadir ground size 64 x 64 km extreme off nadir view 260 x 70 km (approx) rotated 45 deg. Ozone is calculated from the measured radiances by first forming the ratio with measured irradiances (solar view) in each wavelength channel. These ratios, known as directional albedos, are further combined into ratios of directional albedos from 2 wavelength channels. N = -100 Log (A1/A2) Tables of ozone amount vs N-value are precomputed from a variety of atmospheric models using a radiative transfer computer program. For details of this computation and the corrections involved, see Herman et al, "A new self-calibration method applied to TOMS and SBUV backscattered ultraviolet data to determine long-term global ozone change", J. Geophys. Res., 96, 7531-7545, 1991.