Meny

Javascript verkar inte påslaget? - Vissa delar av Lunds universitets webbplats fungerar inte optimalt utan javascript, kontrollera din webbläsares inställningar.
Du är här

Remote Sensing of Forest Decline in the Czech Republic

Publiceringsår: 1998
Språk: Engelska
Sidor: 150
Publikation/Tidskrift/Serie: Meddelande från Lunds Universitets Geografiska Institutioner, Avhandlingar
Volym: 135
Dokumenttyp: Doktorsavhandling
Förlag: Lund University Press, Box 141 S-221 00 Lund, Sweden,

Sammanfattning

This thesis describes the localisation and quantification of deforestation and forest damage in Norway spruce forests in northern Czech Republic using Landsat data.

Severe defoliation increases the spectral reflectance in all wavelength bands, especially in the mid infrared region. These spectral differences allow the separation of three damage categories with an accuracy of 75% using TM data and regression-based relationships. Estimating the same categories using an artificial neural network, multitemporal TM data and topographic data yields slightly higher accuracy (78%). The methods are comparable when using identical input data, but the neural network more efficiently manage large input data sets without pre-processing.

The estimated coniferous deforestation in northern Bohemia from 1972 to 1989 reveals especially affected areas between 600 and 1000 m.a.s.l. and on slopes facing south and southeast. The sector downwind a large source of sulphur dioxide was strongly deforested.

Comparing regional forest damage statistics to three methods estimating harmful effects of sulphur dioxide on Norway spruce yielded significant relationships versus level of forest damage and accumulated salvage felling. Quantifying the effect of data uncertainties permit mapping the probabilities of areas to be significantly over or below thresholds for harmful effects on spruce forests.

Satellite based estimation of coniferous forest health is a good complement to field surveys and aerial photography.

Disputation

1988-05-20
13:15
Geografiska institutionernas föreläsningssal, 3:e våningen, Sölvegatan 13, Lund
  • Curtis Woodcock (Professor)

Nyckelord

  • Earth and Environmental Sciences
  • Czech Republic
  • neural networks
  • sulphur dioxide
  • air pollution
  • Norway spruce
  • deforestation
  • forest decline
  • GIS
  • spectral characteristics
  • Remote sensing
  • Landsat
  • Ore Mountains
  • Geology
  • physical geography
  • Geologi
  • fysisk geografi

Övriga

  • ISSN: 0346-6787
  • ISBN: 91-79-66-528-4
  • ISRN: LUNDBS/NBNG--96/1135--SE

Box 117, 221 00 LUND
Telefon 046-222 00 00 (växel)
Telefax 046-222 47 20
lu [at] lu [dot] se

Fakturaadress: Box 188, 221 00 LUND
Organisationsnummer: 202100-3211
Om webbplatsen