GPP: a program to automate the geophysical data processing

of 6

Please download to get full document.

View again

All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
6 pages
0 downs
ABSTRACT The aim of this paper is to present a program archived for geophysical data processing. The development of the present software package is based on statistical criteria. The GPP package was evaluated by using magnetic, electromagnetic and
  GPP: A Program To Automate The Geophysical Data Processing G. Kalokerinos  Institute of Computer Science. Foundation for Research & Technology - Hellas P.O. Box 1385 Heraklio, GR-711-10 Crete, GREECE,    E. Kokinou and A. Sarris  Laboratory of Geophysical-Satellite Remote Sensing & Archaeo-enviroment, Institute of  Mediterranean Studies. Foundation for Research & Technology - Hellas, P.O. Box 119, Rethymnon74100, Crete, GREECE,  ,  F. Vallianatos   Technological Educational Institute Crete, 3 Romanou Str. Chalepa, Chania, Crete, GR-73133GREECE,  ABSTRACTThe aim of this paper is to present a programarchived for geophysical data processing. Thedevelopment of the present software package isbased on statistical criteria. The GPP packagewas evaluated by using magnetic,electromagnetic and soil resistance datacontaining high levels of noise. The efficiencyof the GPP package was tested thoughcomparison of the results with similarmethodologies followed in a manual procedureusing the Surfer tools.The results archived using the GPP software aremore than satisfactory, enabling the fast andhigh quality production of images even duringthe fieldwork activities.1. INTRODUCTIONThe widespread use of the geophysical researchin archaeological sites (Sarris & Jones, 2000,Jones & Sarris, 2001) resulted in an increasingneed of fast and high quality processing in situ.The existing software packages (such as Surfer,Oasis, Geoplot, etc) provide a suite of tools forthe statistical processing and mapping of thegeophysical data, but they have been provedinefficient in the cases of noisy data. By theterm noisy data is considered the presence of spikes, stripping and zig - zag effects, whichdisturb or even mask the detection of archaeological features. GPP package wasdeveloped in order to encounter these effects bymeans of minimizing the processing time andimproving the resulting image quality of thegeophysical measurements.The data for the evaluation of the GPP packagewere collected in the context of the geophysicalprospection campaigns from Istron, E. Crete(2003 field season) and Dodoni, Epeiros (Sarris,2004).The campaign program at the archaeologicalsite of Istron (Priniatikos Pyrgos) 1 employedground based survey techniques (magnetic, soilresistance and electromagnetic techniques),auger coring for studying the distribution of themagnetic susceptibility, aerial and satelliteimagery for mapping the surface architecturalremnants and GIS analysis. The ultimate goal of the campaign was to retrieve the surface andsubsurface information context of the site andstudy the economic resources of the area inprehistoric and historical periods based on theanalyses of the geomorphological settings andthe regional geology and finds. In 2003 surveyperiod, geophysical survey was expanded to afew vegetable gardens and olive tree fields tothe east, west and southeast of the promontory.In some cases, recent cultivation practices andperiodic flooding has disturbed the upper layersof the site.In the archaeological site of Dodoni,geophysical research was focused around the 1 A multi-disciplinary project implemented through theInstitute for Mediterranean Studies – Foundation of Research and Technology, Hellas (F.O.R.T.H.),Demokritos Archaeometry Laboratory and theMediterranean Section of the University of PennsylvaniaMuseum.  ancient theatre 2 . Magnetic, electromagnetic, soilresistance and GPR techniques covered an areaof 16.000m 2 . One of the goals of the researchwas the detection of the outline of the ancientstadium, which has been partially excavated.The stadium extends to the west of the theatreand was built at the end of the 3 rd century B.C.A subsurface water and electric power network consisting of pipes and cables has heavilydisturbed the upper layers of the area of interestcreating increased noise levels in allgeophysical datasets.A number of different datasets were selectedfrom the above areas in order to test theefficiency of the GPP tools. These included highresolution (0.5m) measurements of the verticalmagnetic gradient with a Geoscan FM36fluxgate gradiometer, soil resistance dataobtained by a Geoscan RM15 resistivity meterwith a Twin Probe configuration of electrodesand electromagnetic measurements acquired bya Geonics EM31 conductivity meter.2. METHODOLOGYOne of the most challenging aspects of our work is addressing methodological issues inimproving the processing time and the dataquality.The GPP package was developed on a LINUXplatform by using GCC compiler and then wasported using Borland C compiler in order to beexecuted in a command (GCC) window inWINDOW NT environment.The implementation of the GPP package issequential. The grid size may vary from a fewdozen of points to a few thousands points. As aninput it accepts geophysical data in an ASCIIformat (X,Y,Z columns) and then a series of interactive statistical processes is followed.In the next paragraphs we describe the basicoperations of the GPP package. Preprocessing In general, the GPP package runsby creating a file list of the files to be processed.In case there is a need of reversing the X and Y 2 The geophysical campaign at Dodoni was carried out bythe Laboratory of Geophysical-Satellite Remote Sensingand Archaeo-environment (IMS-FORTH) under theauspices of the Committee for the Protection andRestoration of Monuments of the Archaeological Site of Dodoni. columns (in order to bring grids in the rightalignment with respect to the north), thePreprocessing option (Fig. 1) is chosen.Thereinafter the geometry of the grid is checkedand Mean and Standard Deviation values arecomputed. Maximum and minimum values of the measurements, Mean, Standard deviation, Xand Y coordinates and X and Y samplinginterval are displayed. At his phase of theprocessing peak values can be muted and thebenchmark relocated by shifting the X, Ycoordinates. The preprocessed file is saved as*_PR.dat file.  Main Processing constitutes the core phase of the processing and manipulation of the data.The Main Processing procedure (Fig. 2) isdivided in three steps. In the first step (Levelcorrection) the dynamic range of all files isspecified according to the Grid Level Correctionfactor. Thereinafter the de-spiking techniquecan be applied according to the noise level. Incase of noisy data the De-spiking Factor mustbe smaller than one standard deviation in orderto smooth the data. In case of less noisy data theDe-spiking Factor can be greater than onestandard deviation. Finally, Line Equalization isapplied to the data in order to avoid strippingeffects.The Level corrected file is saved as *_L0.datfile, while the Level corrected and de-spiked fileas *_L1.dat file. The Level corrected, De-spikedand Line equalized file is saved as *_L2.dat.Preprocessing and Processing reports aregenerated by the end of each procedure. Secondary options   of the GPP package includere-sampling of the data by any step in X and Ydirection and exclusion of the dummy values.GPP is also able to make a mosaic of grids atany stage of processing (preprocessed and mainprocessed grids).3. EVALUATION OF THE GPP PACKAGESelected geophysical data collected from thearchaeological sites of Istron (PriniatikosPyrgos) and Dodoni were chosen in order to testthe validity and the efficiency of the GPProutines. Most of the selected datasets sufferedfrom increased levels of noise.Magnetic data were collected by 7 grids locatedSE of Priniatikos Pyrgos (Fig. 3). The area liesin the vicinity of a high voltage power supply  pillar. Concentration of surface shards suggeststhe past occupation of the site during the Romanperiod. Original data suffered from a leveldifference between the individual grids. Anelevation difference between the southern andthe northern sections of the sites has alsoimposed problems in the connection among thegeophysical grids. Furthermore, a linear trend ispresent within a few grids. Processing by meansof the GPP tools managed to eliminate theinternal differences among the grids andproduce a uniform mosaic, which outlines anumber of features. Two parallel anomaliesalong the NW-SE directions (along the northernand southern sections) meet a perpendicularlinear anomaly to the SE suggesting theexistence of a possible structure of dimensionsabout 30x30m.In another field close to the coast (east of thePriniatikos Pyrgos promontory) (Fig. 4),geophysical data collected by magnetic,resistivity and EM techniques exhibited extremelevel of noise due to recent clearance andconstruction activities. A trend towards lowresistivity values is shown to the northernsection of the soil resistance data, probablysrcinating by the proximity of the grids to thecoast line. De-spiking techniques, Lineequalization and Re-sampling in the spatialdomain removed the most intensive anomaliescaused by non - archaeological features. Theresulting anomalies to the south section of thearea could be correlated to surface architecturalremnants which are obvious to the south.In a different situation, both magnetic (Fig. 5a,)and electromagnetic (Fig. 5c) measurementsacquired west of the ancient stadium in Dodoniwere masked by the intensive anomaliessrcinating by water pipes and electrical cableslocated within a meter below the currentsurface. Extreme values caused by the existenceof widely lain debris and metal fragments andlinear and non-linear instrumental drift in bothdirections made processing of the data adifficult task. The resulting maps following theprocessing steps of GPP (Figs. 5b & 5d),indicate a drastic improvement of the dataquality, which made it possible to recognizesome subtle anomalies to the SW section of thearea, probably related to a cluster of architectural relics.4. CONCLUSIONSGPP utilities proved satisfactory for the type of enhancement required by the geophysical dataresulting from the shallow depth prospection of archaeological sites. Interactive de-spiking bymeans of statistical processing of the srcinalmeasurements seems to be a flexible techniquein order to remove peak values that maskedspecific areas in their vicinity, while lineequalization (in one or two directions) smoothesthe grid and makes easier the detection of archaeological features. Finally, re-samplingtechniques comprise a useful tool in removingstripping effects.The full processing of the geophysical data byGPP package in situ can be proved a useful toolto the interpreters and research designers.REFERENCES Sarris, A. & Jones, R., 2000, “Geophysical and RelatedTechniques Applied to Archaeological Survey in theMediterranean: A Review”,  Journal of Mediterranean Archaeology (JMA), v.13, no.1, pp. 3-75, June 2000.Jones, R., & Sarris, A., 2001, Electronic Databanks forMediterranean Geophysical Surveys: Expectations andDeliverables for Practitioners & Archaeologists, 4 th Int.Conf. On Archaeological Prospection, Austrian Academyof Sciences, Vienna, September 19-23, 2001.Sarris, A., 2004, Technical Report of the Work of theGeophysical Research in the Archaeological Site of Dodoni, unpublished report, Institute for MediterraneanStudies – FORTH, Rethymno, 27 March 2004.   Figure 1. Procedure of the Preprocessing option.Figure 2. Procedure of the main processingoption.    0 10 20 30 40 500102030405060 -7-5-3-11357 nT/MnT/M 0 10 20 30 40 500102030405060 -45-30-15015304560 nT/M 0 10 20 30 40 500102030405060 -3.5-2.5-1.5- (a)(b)(c)  Figure 3: Raw magnetic data (Vertical magneticgradient, nT/m) collected in the area of Istron, (b)Manually processed magnetic data by using Surfer tools,(b) Automatic processing of the magnetic data by usingGPP package. Ohms 0 5 10 15 20 25 30051015202530354045 -5515253545556575 0 5 10 15 20 25 30051015202530354045 61116212631 OhmsOhms 0 5 10 15 20 25 30051015202530354045 020406080100 (a)(b)(c)  Figure 4: (a) Raw soil resistance data (Ohms)collected in the area of Istron, (b) Manuallyprocessed data by using Surfer tools, (b)Automatic processing of the magnetic databy using GPP package.
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks