Bucharest Plates and CCD Database Petre Popescu, Gheorghe Bocsa, Alin Nedelcu, Liviu Serbanescu, Octavian Badescu Astronomical Institute of Romanian Academy.

Bucharest Plates and CCD Database

Petre Popescu, Gheorghe Bocsa, Alin Nedelcu, Liviu Serbanescu, Octavian

Badescu

Astronomical Institute of Romanian Academy5 Cutitul de Argint street, Bucharest, Romaniawww:astro.ro

1. Introduction

Within the Wide-Field Plate Archive program initiated by IAU Commission 9,

• 2260 plates were archived in the period 1930 – 1951 at Bucharest Observatory – under the aegis of University;

• 10729 plates were exposed from 1952 until now, when the Observatory was affiliated to the Institute of Atomic Physics and the Romanian Academy.The observations were performed in the zone – 200<δ<+700 (English mounting) with a Prin Merz refractor (f=6m, d=38cm) and Zeiss Canon camera (f=80cm, d=16cm). The refractor field is up to 20×20 (1mm=30’’) and that of Zeiss Canon camera is 70×100 (1mm=4’). Zeiss camera was used only for the exposure of 180 photographic plates of 13×18cm.

2. Qualitative study of the plates observed with Bucharest astrograph

The observational program had • 3 maxima: 1930-1940, 1952-1977 and 1980-1998;• 3 minima: Second World War, 1977 Earthquake (the dome was

damaged), 1998 - deteriorated astroclimatic conditions. In the first period, observations were made for asteroids, as well as for some great planets, comets and lunar eclipses. In spite of the long time elapsed most plates have preserved a good aspect. While finalizing the “previews” for the first group of 2260 plates, we have noticed that the content of many plates is not the expected one because the observers’ lack of professionalism and collaboration. In that time, the Observatory didn’t issue any publication, processing was difficult and lacked calculations means, (besides we suspect) that few processing were sent to Rechen Institute in Berlin. There were performed observations in the hope that in the future processing might also be performed.


• In the first period, good plates with adequate exposure is a valuable source of asteroids and comets positions they indicate the orbits of that bodies.

• An activity acknowledged worldwide was the observation of Jupiter’s Galilean satellites by prof. Gh. Petrescu, who built a mechanic device by means of which he was able to photograph with adequate exposure times the reference stars, satellites: Io, Europe, Ganymede, Callisto and Jupiter! His work made up a doctorate thesis which was defended at Paris Observatory – all data being published, also where we suspect that the plates are. The observations were resumed in the period 1982-1985 and new results were published.


• A special event was the discovery in the archive of a number of plates with Pluto, observed by the academician Gh. Demetrescu in January 1932, together with other observations performed in 1975 by Cornelia Cristescu – processed with respect to the epoch J2000.

• In 2001 the paper “Observations of Pluto in Bucharest during 1932 and 1967-1975 ” was published, data processing being made for the epoch J2000.

• In the plate archive we have found several high quality plates with Daimaca comet from September 1943.


• The second period brought the acknowledgement of photographic astrometry working group. - There were performed comet and asteroid observations to improve their orbits, the data were finalized in Smithonian Astronomical Observatory; - Observations of great planets: Jupiter, Neptune and Pluto;- Photographic determinations of the orbits of Galilean Satellites of Jupiter – Petrescu (1935, 1938, 1939) and Bocsa (1994);- In 2009 Neptune will make the first complete rotation around the Sun since its discovery by Le Verrier and Adams in 1846. The paper “Observations of Neptune in Bucharest Observatory during 2001-2004” was published in Romanian Astronomical Journal to this effect.


• Because in that period was no telescope in Bucharest, we use the astrograph for observations of stellar clusters (24 plates), double stars (81 plates) and variable stars (105 plates).

Within the framework of this program was observed the Pleiades region for the period 1955-1958. It was possible to use them for the project investigation of the long variability of the Pleiades red dwarf stars, together with Byurakan, Konkoly, Rozhen, Postdam and Sonnenberg Observatories. There are more than 80 plates, which might be object of future usage of the archived observations due to the long period time coverage.


Great programs collaboration in the field of photographic astrometry: • Between 1957 – 2000 we participated in the program initiated by

Rechen Institute Heidelberg and Institute of Theoretical Astronomy Leningrad for observations of a designated number of asteroids, as well as for the determination of an inertial system of stellar reference.

• Within “Stellar proper motions” program with Pulkovo Observatory, at the first stage (1957 – 1965) we used 151 galaxies and 275 fundamental stars as fixed reference points for the determination of stellar proper motions, zone - 200<δ<+700. At the second stage (1980 – 1990), faint distant stars were used. During this stage, because of the deterioration of the astroclimatic conditions in Bucharest, only stellar areas with 14m were used, in the zone - 200<δ<00.


• The program “CONFOR” (started in 1992) has as aim to make the connection between radio and optical reference frames. In the first step, we observed 188 stellar areas with a declination range of - 200<δ<+700 (finalized in 2000). In the second step (which is in work now) we use CCD techniques to observe the radio-sources and the reference system around them. This step is performed in Belogradchik Observatory using a 60cm Zeiss reflector (see part 4).

• ESA program “COROT” envisages the observational analyze of two zones, by comparing old photographic plates with new exposed plates, for the evaluation of the dynamics of this two stars in the Galaxy. For the first star we found a pair plate of 13×18cm in the year 1939. For the second in WFPDB we found in Sonnenberg Observatory a plate observed in 1950 (6h 48m, -00 30m).


• Occasionally observations of phenomena as: total or partial solar eclipses, Regulus occultation by Venus (1959), Mercury transits across solar disk (1970, 1973).

• Photographic material is still in good shape due to the good preservation conditions in the plate archive. There are several plates which have not been washed properly, but when the previews will be done, they will be reconditioned. The plates 13×18cm, made with Zeiss camera are underexposed and for many plates from the first stage (1930 – 1951) the observational subjects cannot be found in the images obtained. The content of the plates after 1980 is good, but because sky conditions they are darker and increasingly difficult to process. Till now we have 2300 preview plates (1930 – 1951), in the end of 2008 we will have 10000 preview plates (1930 – 1980).

3. Plate information recovery and analyse

In the digitized information from the plates archive may occur errors caused by several factors as:

- physical degradation of the plates,

- the digitization process (from scanning, photographing, etc.),

- measurement processes (to obtain the coordinates for certain celestial objects that are on that plate), etc.


Comparing common areas with similar measuring methods, it is possible to determine models for the distribution of the errors (for errors that may occur during digitization process, respectively the measurement process) or evolution models of the errors (in the case of degradation of the plates).

These models may be used to compensate the occurring errors or to delimit them. In a digitized database the comparing operations for common areas may be done automatically by the software. It will be obtained the determination of the identical areas from star catalogs with minimal tolerance from the measurement's and classification's point of view, having as result fuzzy models for the occurring errors.

Using these fuzzy models the digitized plates library may give information regarding the level of "trust" level of each digitized image.


• Clasification of the errors The errors that occur, from the digitized information of the plate archive, can be divided in errors that do not change in time – static errors – and errors that change in time – dynamic errors.- A. Dynamic errors - due to physical degradation of the plates (mainly defect of the photographic emulsion with evolution in time, materialized by the change of magnitude on the plates);- B. Static errors - measuring errors caused by: - B.1. Equipments used for digitization; - B.2. Methods or softwares used during the measuring process; - B3. Other causes – difficult to be quantified (results of the photographing process, initial defects of plates, etc).


• Error's detection and model’s construction 1. Error's detection by software (in a database which contains digitized plates)It is possible to make a software application which analyze data in a databases (with digitized plates) and to determine the common areas (areas that will be found in more then one digitized plate - from same or different observatories). These detected areas by the software are entering in the error's analysis process. Based on information that are corresponding to selected digitized plates it can be built fuzzy models.In this models may enter, effectively, all associated information of the digitized plates.


• Error's detection and models construction

2. Error's detection by experiments (Realized on samples chosen after series of classification in plate archive)A. To put in evidence the resulted errors from time degradation of plates may be done by successive digitization of same plate. It can be performed at few months interval and making measurements for the references of the plate in the same conditions. From deviations, will be build an evolution model in time. Applying regression techniques we will determine the mathematical model. The model will be in one of the following form:

Ym = Fm(p,m,t) where Ym represents the deviation of the magnitude, p - depends of the type of the plate,m - depends of the measurement method,t - is the time (the number of months)


• Error's detection and models construction

2. Error's detection by experiments (Realized on samples chosen after series of classification in plate archive)

B.1. The errors that are generated by the models or by the software used in the measuring process.

The measurement may be evidenced measuring the reference celestial object position for the same digitized plate using different methods and software.


The models will have the following form:

Ys = Fs(m,r,i,p,s) Yu = Fu(m,r,i,p,s) Ya = Fa(m,r,i,p,s)

where Ys represents the deviation from the magnitude,

Yu represents the angle where the deviation appeared,

Ya represents the absolute value of the position's deviation, s represents the type of the scanner,r represents the the scanning resolution, i represents the pixel from the imagep - depends on the type of platem - depends on the measurement method,The model will be a fuzzy model.


B.2. The errors due of the used equipments for digitization, may be put into evidence digitizing same plates with different equipments, using the same measuring methods to determine the position of the reference celestial objects. The models would have the following forms: Ys = Gs(s,r,i,p,m) Yu = Gu(s,r,i,p,m) Ya = Ga(s,r,i,p,m) ,

• where Ys represents the deviation from the magnitude, • Yu represents the angle where the deviation appeared,• Ya represents the absolute value of the position's deviation, • m is a measure that depends by the measuring method and the

measuring software,• s represents the type of the scanner,• r represents the the scanning resolution, • i represents the pixel from the image• p is a measure that depends on the type of plate.• The model will be a fuzzy model.• The software may replace the experimental test to detect and create

models for digitized plates if the digitized plates archive already exists.


A. The errors models obtained by analyze of dynamic errors (degradation in time of the plates) is useful for:- setting the priorities regarding plates scanning;- for some plates, the old ones, may improve the measurements (the corrections applied based on the model will bring the digitized image closer to the initial one).B. Models resulted from the analyze of static errors may be used for the improvement of the quality of the measurement techniques as well as for choosing the optimal equipments for digitization.- The first benefit of the database brought by the software is linked to the “trust level” that may be attached to each plate that contains common areas with other plates that exist in the digitized plates archive.- In addition, the software will offer errors’ analyzing models corresponding to the plates.

4. Database of FITS images of extragalactic radiosources

•Astronomers are able to share data (images, spectra), because they share a common data description called FITS (Flexible Image Transport System).

•The FITS rules are endorsed by the International Astronomical Union, and codified into a formal standard by the NASA/Science Office of Standards and Technology.

•The FITS standards make observations from any instrument usable by any astronomer, and many generic tools for data visualisation and transformation have been developed.

•Various programming libraries provide the infrastructure of the input/output high level function needed to access the FITS data – QFITS (ESO), CFITSIO (NASA)


The CCD images of the extragalactic radiosources observed so far are stored on DVDs.

The total amount of data is 26 GB including the science and calibration images. For an easy access to these images obtained in the last 3 years the FITS structure is stored in a MySQL database

An web interface based on PHP provides a reasonably flexible way to query the database A more powerfull way to access the database is to use SQL queries in stand-alone C programs running in batch mode


•The database is updated in remote mode after every observational run;

•A stand alone C program using CFITSIO library to access the FITS file and MySQLCLIENT library to communicate with the database is parsing the folder of the last observations;

•FITS headers togheter with the actual path of each image are stored in the MySQL database;

•SQL query example: find all the regions observed at a given date (JD): SELECT RA, DEC from DATABASE where TJD >JD -1 AND TJD < JD + 1.

5. Conclusion

•Reusing the data from older projects (photographic plates) for new scientific objectives is also necessary to optimise their scientific return (asteroids recovery and/or discoveries);

•The need to uniformally describe the heterogenous data (tables, images), or data of similar types but obtained on different arhitectures (photographic plates, CCD images);

•The perspective of a European regional storage center and a colaborative data-mining project;

•New methods, tools, and techniques for the preparation and use of the existing data.

Bucharest Plates and CCD Database Petre Popescu, Gheorghe Bocsa, Alin Nedelcu, Liviu Serbanescu, Octavian Badescu Astronomical Institute of Romanian Academy.

Documents