Wyniki wyszukiwania - BazTech

Ograniczanie wyników

1 2005

Powiadomienia systemowe

Sesja wygasła!

Znaleziono wyników: 1

Liczba wyników na stronie

Wyniki wyszukiwania

Wyszukiwano:
w słowach kluczowych: strefy uszkodzeń lasu

Sortuj według:

Ogranicz wyniki do:

Rola klasyfikacji nadzorowanej wyzokorozdzielczych obrazów satelitarnych QuickBird w nowej koncepcji wyznaczania przemysłowych stref uszkodzeń drzewostanów na przykładzie Miasteczka Śląskiego

Kozioł K., Wężyk P.

Roczniki Geomatyki

2005

T. 3, z. 2

87-96

During the last decades the influence of industry left its trace on Polish forests worsening significantly their health condition, expressed e.g. by the defoliation degree. Direct negative influence of industrial contamination was in certain forests so high that they have been considered the zones of ecological disaster. The new Instruction of Forest Management issued in 2003 does not describe precisely the method of zone delineation . leaving a serious gap . due to which the zones determined 10 or 20 years ago are taken over by new forest management. The purpose of this paper is to point out the usefulness of application of the VHR QuickBird satellite images (04.09.2004; region of 89 km2), which is the socalled supervised classification in assisting the new method of the forest damage zones delineation. Forests of the .wierklaniec Forest Inspectorate (RDLP Katowice) surrounding the zinc and lead metallurgical plant in Miasteczko Śląskie was the region of interest. Distinct damages of stands, including even the so-called industrial turfs caused by heavy metals dust emission and toxic concentrations of gases are still visible in this region. The calibration of image was based on vector data from the Digital Forest Map and GCP (RMSXY <1.50 m) control measurements. The supervised classification (Maximum Likelihood) was performed by means of DGPS measurements in ROI. The results of classification of the QuickBird image were compared with the determined Test Areas and information in the database of the LAS/SILP system prepared during the last forest management review (2003). The obtained classification likelihood for the .FOREST. class was 92.6%, Chile for the .NON FOREST. class: 95.3%. The likelihood of the classification for individual tree species was from 82% for Scots pine (Pinus sylvestris L.) to 38% for Black alder (Alnus glutinosa L.). For industrial damage zones (II, III and IV) determined more than 10 years ago, the supervised classification allowed to establish the approximate parameter of the crown density as being: 90.37% (zone II), 90.43% (zone III) and 71.93% (zone IV . the so-called .death zone.). Detailed analyses of the selected evaluating features proved the possibility of utilizing the QuickBird image for the determination of such parameters as the crown density or the number of trees in a unit area. The value of the crown canopy oscillated from 94.9% to approximately 7.9% in the damage zone IV. The investigations have shown, beyond any doubt, that the VHR QuickBird satellite images can assist works on site and decision-making processes in delineation of new damage zones. High resolution in the region of interest and variety of radiometric information cause that the QuickBird images are valuable cartometric material, which due to the integration with other geoinformatic technologies (GPS, GIS) can be used for the verification and updating of LMN and SILP databases, including information on already existing forest damage zones.