Topographic Map Construction Using a Handheld Laser Range Finder and GIS at Kaman-Kalehöyük and Kültepe

Transcription

1 Topographic Map Construction Using a Hanhel Laser Range Finer an GIS at Kaman-Kalehöyük an Kültepe Yuichi S. HAYAKAWA an Kaoru KASHIMA Tokyo Fukuoka ABSTRACT Detaile topographic maps for areas aroun Kaman-Kalehöyük an Kültepe were constructe using a hanhel laser range finer an GIS. The equipment for laser measurement enables us to make a fine topographic map with less time an working effort compare to traitional measurement methos. Keywors: Lanform, Laser measurement system, DEM, GIS, Kriging INTRODUCTION Topographic maps are funamental units of information for many types of fiel stuies, incluing geosciences an archaeology. Although various surveying methos exist to measure lan surface topographies, they are sometimes problematic in terms of time an money. Traitional methos of making contour-base topographic maps, such as plane-table surveying, total stations an GPS, often require heavy, large instruments an several operators, as well as a long time to complete fiel measurements. Recently evelope surveying approaches such as aerial/satellite-image photogrammetry an airborne/terrestrial laser scanning systems allow one to obtain fine topographic contours, DEMs (Digital Elevation Moels) an/or three imensional surface ata of the objects, but those high-technological measurement systems are usually too expensive to be casually use. To obtain a topographic map of a certain area of hunres of square meters in an aroun the excavation sites of Kaman-Kalehöyük an Kültepe, we performe a quick an cost-effective survey of lanform measurement using a laser range finer, following the methoology propose by Hayakawa et al. (in press). MEASUREMENT DEVICE The laser range finer use in this stuy is LaserAce 300 by Measurement Devices Lt., UK, which can measure range, vertical angle an horizontal angle. The evice has a size of mm an a weight of 650 g, which is small an light enough to carry easily. An operator looks at the target object through its viewfiner an a laser beam (wavelength: 905 nm; laser safety: Class I) is emitte onto the object when a button is pushe. A sensor in front of the evice then catches the light reflecte from the target object within 0.3 secons. The range ata are available without a target reflector, but they are often unavailable for some specific materials such as water surfaces an ark or rough materials which cannot reflect the laser appropriately. The measurable range is up to 300 m, with an accuracy of ±0.1 m an a resolution of 0.01 m. The maximum measurable istance changes epening on the atmospheric conition: the maximum istance generally ecreases as the atmospheric humiity or ust ensity increases. Together with istance, an elevation angle an a horizontal angle are concurrently obtaine. The range of measurable vertical angle is ±80 with an accuracy of 0.3 an a resolution of 0.1. The horizontal

2 192 Y. S. HAYAKAWA an K. KASHIMA AAS XV angle, of which accuracy is less than 1 an resolution is 0.1, is expresse as an angle from magnetic north. The resultant istance, vertical angle an horizontal angle are isplaye on the winow of the evice at each measurement operation. at the altitue m a.s.l., locate in the excavation area of the Karum. The ata set, covering an area of approximately 800 m 1,200 m, inclues the current excavation area of the Karum, but not the moun of Kültepe / Kanis. DATA ACQUISITION The measurement evice was hanhel by a staning operator. The vertical offset, i.e., the height of the operator s eyes, was subtracte from the original ata in the ata conversion processes. Horizontal istance, vertical height an horizontal angle from the evice to the target point were manually logge. Measurements were carrie out base on multiple control points. The first control point (the base point) was set at a istinct location in the excavation area, an its longitue (x), latitue (y) an altitue (z) were measure with the GPS. Other control points were then successively measure in areas where one can walk in. From each control point, target points on terrain surfaces were ranomly measure. For areas covere with vegetation such as sunflower fiels, the surfaces of the vegetation were first measure an the height of the vegetation (sunflower or other bushes) was subtracte uring the ata conversion processes later. Aitional information about lanmarks such as roas, electric poles an river banks was put into the ata sheet an later use to ientify the objects an to create their vector ata. Kaman-Kalehöyük The base point was set at the southwest corner of the excavation trench on the tepe at longitue E, latitue N an altitue m a.s.l. During the three ays of our survey, the ata of 2,049 points were measure with 42 control points. The ata are mainly from the eastern sie of the tepe an the ajacent hill slopes, covering an area of approximately 700 m 1,100 m. Kültepe In orer to cover a larger area in a shorter perio than Kaman, we collecte a total of 404 points with 45 control points in two ays at Kültepe. The coorinates of the base point were E an N, MAP CONSTRUCTION The fiel ata were entere into a personal computer to hanle them in a igital format. Three-imensional coorinates (x, y an z) were calculate from the istance, height an horizontal angle ata. The xyz ata an the aitional information of the local features (roas, electric poles an trench outlines) were then importe into GIS software (ESRI ArcView) as point ata. The coorinate of the base point was set in the global coorinate system. The longitue an latitue of the base point were projecte on UTM coorinates, permitting the measure metric ata to be put onto the map with other global ata sets in the UTM projection. For Kaman, the coorinate of the base point is ( , ) in UTM Zone 36, while the base point coorinate for Kültepe is ( , ) in UTM Zone 37. Lan surface topography ata (DEM) were create by interpolating the z-values of the obtaine point ata on the GIS. The interpolation was carrie out by Kriging (e.g., Burgess an Webster 1980), which has been wiely applie to estimate spatial istribution of various phenomena from scattere point ata. The resolution of the DEM was etermine by the ensity of the points calculate for each m gri cell. The maximum ensity is m -2 for Kaman an m -2 for Kültepe. These ensity values inicate that the minimum of the mean istance between two ajacent points is 10.5 m for Kaman an 16.0 m for Kültepe. The gri resolutions (cell size) for Kaman an Kültepe were thus etermine to be 10 m an 20 m, respectively. Testing the semi-variograms, the Kriging interpolation was carrie out using the circular function with no nugget effect, for which, 12 ajacent points within a istance of 300 m were use. Some errors were observe as anomalous z-values

3 2006 Topographic Map Construction 193 in the ata set of Kaman, which inclue more than 2,000 points. These coul be erive from protruing branches of the vegetation an/or the misalignment of the control points resulting from error propagation through the successive control points. The error points were exclue by looking into the tentative 1-m contour lines generate from the original point ata set before construction of the final DEM an contour ata. Anomalous points (spikes an sinks) more than 2 m higher or lower than surrouning points can be ientifie as points surroune by two or more circular contour lines. Eliminating those anomalous points, interpolation was again performe over the remaining point ata set. These proceures were repeate several times until most of the error ata were iminishe. Consequently, 2,035 points remaine from the total 2,049 points. From the interpolate DEM, topographic contour lines with 1-m interval were erive. Vector ata (polygons, polylines) of the groun object features (builings, roas, excavation trench outlines an riverbank lines) were also etermine with the aitional information attribute to the point ata. Overlaying the layers of the contour lines an other vector feature ata, high-resolution topographic maps for Kaman an Kültepe were obtaine (Figs. 1, 2). Comparing the resultant map an the actual lanscape, the istance accuracy seems to be within several meters. The vertical accuracy may also be within one to two meters, because the prominent errors eliminate uring the error reuction stages were no taller than three meters. However, it is somewhat har to fully assess the accuracy of the map because of the lack of accurate comparable ata. More fiel surveys for further ata collection will be necessary to obtain the topography ata in a wier area an in more etail. BIBLIOGRAPHY Hayakawa, Y. S., T. Oguchi, J. Komatsubara, K. Ito, K. Hori an Y. Nishiaki Rapi on-site topographicmapping with a hanhel laser range finer for a geoarchaeological survey in Syria, Geographical Research (in press). Burgess, T. M. an R. Webster 1980 Optimal interpolation an isarithmic mapping of soil properties. 1. The semi-variogram an punctual Kriging, Journal of Soil Science 31 (2), pp Yuichi S. HAYAKAWA Department of Earth an Planetary Science Grauate School of Science The University of Tokyo Hongo, Tokyo Japan hayakawa@csis.u-tokyo.ac.jp Kaoru KASHIMA Department of Earth an Planetary Sciences Kyushu University , Hakozaki, Fukuoka Japan kashima@geo.kyushu-u.ac.jp

4 Y. S. HAYAKAWA an K. KASHIMA AAS XV N W E S Fig. 1 Topographic map on the eastern sie of Kaman-Kalehöyük. Contour interval is 1 m.

5 Topographic Map Construction Fig. 2 Topographic map aroun the Karum of Kültepe.

6