- High Spatial Resolution Remote Sensing Data for Forest Ecosystem Classification: An Examination of Spatial Scale
Treitz, P; Howarth, P
Remote Sensing of Environment [Remote Sens. Environ.], vol. 72, no. 3, pp. 268-289, Jun 2000
Detailed forest ecosystem classifications have been developed for large regions of northern Ontario, Canada. These ecosystem classifications provide tools for ecosystem management that constitute part of a larger goal of integrated management of forest ecosystems for long-term sustainability. These classification systems provide detailed stand-level characterization of forest ecosystems at a local level. However, for ecological approaches to forest management to become widely accepted by forest managers, and for these tools to be widely used, methods must be developed to characterize and map or model ecosystem classes at landscape scales for large regions. In this study, the site-specific Northwestern Ontario Forest Ecosystem Classification (NWO FEC) was adapted to provide a landscape-scale (1:20 000)forest ecosystem classification for the Rinker Lake Study Area located in the boreal forest north of Thunder Bay, Ontario. High spatial resolution remote sensing data were collected using the Compact Airborne Spectrographic Imager (CASI) and analyzed using geostatistical techniques to obtain an understanding of the nature of the spatial dependence of spectral reflectance for selected forest ecosystems at high spatial resolutions. Based on these analyses it was determined that an optimal size of support for characterizing forest ecosystems (i.e., optimal spatial resolution), as estimated by the mean ranges of a series of experimental variograms, differs based on (i) wavelength, (ii)forest ecosystem class, and (iii) mean maximum canopy diameter (MMCD). In addition, maximum semivariance as estimated from the sills of the experimental variograms increased with density of understory.
- Accuracy comparison of various remote sensing data sources in the retrieval of forest stand attributes
Hyyppa, J*; Hyyppa, H; Inkinen, M; Engdahl, M; Linko, S; Zhu, YH
Forest Ecology and Management [For. Ecol. Manage.], vol. 128, no. 1-2, pp. 109-120, 15 Mar 2000
Recent advances in developing new airborne instruments and space-borne missions and in SAR technology, especially in interferometry and coherence estimation, have roused questions: can such new SAR data be utilized in operational forest inventory? What is the accuracy of different satellite data for forest inventory? This paper verifies the explanatory power and information contents of several remote sensing data sources on the retrieval of stem volume, basal area, and mean height, utilizing the following data: Landsat TM, Spot PAN and XS, ERS-1/2 PRI and SLC (coherence estimation), airborne data from imaging spectrometer AISA, radar-derived forest canopy profiles (obtained with HUTSCAT), and aerial photographs. Ground truth data included three different sets ranging from conventional forest inventory data to intensive field checking where one man-day was spent for assessing one stand. Multivariate and neural network methods were applied in data analysis. The results suggested that (1) radar-derived stand profiles obtained with 100 m spacing was the most accurate data source in this comparison and was of equivalent accuracy with conventional forest inventory for mean height and stem volume estimation, (2) aerial photographs (scale 1 : 20,000) gave comparable results with the imaging spectrometer AISA, (3) the satellite images used for the estimation in the decreasing explanation power were Spot XS, Spot PAN, Landsat TM, ERS SAR coherence, JERS SAR intensity images (PRI); and ERS SAR intensity images (PRI). It appears that optical images still include more information for forest inventory than radar images, (4) from all satellite radar methods, the coherence technique seemed to be superior to other methods.
- Monitoring forest degradation in tropical regions by remote sensing: some methodological issues
Global Ecology and Biogeography [Global Ecol. Biogeogr.], vol. 8, no. 3-4, pp. 191-198, Jul 1999
Key issues related to the monitoring by remote sensing of open forest degradation in a tropical context are discussed. Degradation of forest-cover is often a complex process, with some degree of ecological reversibility and a strong interaction with climatic fluctuations. Only a representation of land cover as a continuous field of several biophysical variables can lead to an accurate detection of forest degradation. For this purpose, repetitive measurements of spectral, spatial and temporal indicators of the land surface have to be performed. Each set of indicators brings a specific type of information on the land cover. These indicators must therefore be combined to achieve a comprehensive description of the surface processes. The detection of inter-annual changes in landscape spatial structure is more likely to reveal long term and long lasting land-cover changes, while spectral indicators are more sensitive to fluctuations in primary productivity associated with climatic fluctuations. Different monitoring systems may be optimal for different ecosystems. A long time series of observations is always required. The monitoring of the spatio-temporal distribution of biomass burning may also give indications of open forest degradation.
- Remote sensing of forest fires in boreal ecosystem from space
Li, Z; Fraser, R; Khananian, A
PROC SPIE INT SOC OPT ENG, SOCIETY OF PHOTO-OPTICAL INSTRUMENTATION ENGINEERS, BELLINGHAM, WA, (USA), 1999, vol. 3868, pp. 228-231,
This paper presents a comprehensive investigation of Canadian boreal forest fires using satellite measurements. Algorithms were developed for detecting active fires (hotspots), burned areas, and smoke plumes using single-day NOAA-AVHRR images and 10-day AVHRR NDVI composites. The algorithms were rigorously validated using conventional fire survey data. The hotspot algorithm identified almost all fire events, but cumulative hotspot area was significantly smaller (approximately 30%) than burned area reported by fire agencies. The hybrid, burn mapping technique provided estimates of Canada-wide burned area that were within 5 percent of official statistics. A neural-network classifier was also developed that allows smoke plumes to be effectively separated from cloud cover at a regional scale.
- Passive and active microwave remote sensing of soil moisture under a forest canopy
Chauhan, N; Le Vine, D; Lang, R
DIG INT GEOSCI REMOTE SENS SYMP (IGARSS), IEEE, PISCATAWAY, NJ, (USA), 1999, vol. 4, pp. 1914-1916,
Active and passive microwave remote sensing of soil moisture is compared at L-band for forested covered terrain. Data collected in Howland, Maine during the Forest Ecosystem Dynamics (FED) experiment in 1990 have been used to validate a discrete scatter model. The model is then used in a series of parameter studies to extrapolate the results to canopies with a wide range of biomass and surface conditions. The objective of this study is to gain insight into the range of biomass over which soil moisture can be measured by the two sensors. Tree and surface parameters are varied to generate different biomass and ground conditions. Both radar and radiometer exhibit sensitivity to soil moisture under a wide range of conditions. How these sensitivities diminish with increasing biomass and surface roughness is an important factor in the response of both sensors.
- Forest biodiversity and its assessment by remote sensing
Innes, JL; Koch, B
Global Ecology and Biogeography Letters [Global Ecol. Biogeogr. Lett.], vol. 7, no. 6, pp. 397-419, Nov 1998
Several international conventions and agreements have stressed the importance of the assessment of forest biodiversity. However, the methods by which such assessments can be made remain unclear. Remote sensing represents an important tool for looking at ecosystem diversity and various structural aspects of individual ecosystems. It provides a means to make assessments across several different spatial scales, and is also critical for assessments of changes in ecosystem pattern over time. Many different forms of remote sensing are available. While lately the emphasis on laser scanner and synthetic aperture radar data has increased, most work to date has used photographs and digital optical imagery, primarily from airborne and spaceborne platforms. These provide the opportunity to assess different phenomena from the landscape to the stand scale. Remote sensing provides the most efficient tool available for determining landscape-scale elements of forest biodiversity, such as the relative proportion of matrix and patches and their physical arrangement. At intermediate scales, remote sensing provides an ideal tool for evaluating the presence of corridors and the nature of edges. At the stand scale, remote sensing technologies are likely to deliver an increasing amount of information about the structural attributes of forest stands, such as the nature of the canopy surface, the presence of layering within the canopy and presence of (very) coarse woody debris on the forest floor. Given the rate of development in the technology, even greater usage is likely in the future.
- Determining forest species composition using high spectral resolution remote sensing data
Martin, ME; Newman, SD; Aber, JD; Congalton, RG
Remote Sensing of Environment [Remote Sens. Environ.], vol. 65, no. 3, pp. 249-254, Sep 1998
Airborne hyperspectral data were analyzed for the classification of 11 forest cover types, including pure and mixed stands of deciduous and conifer species. Selected bands from first difference reflectance spectra were used to determine cover type at the Harvard Forest using a maximum likelihood algorithm assigning all pixels in the image into one of the 11 categories. This approach combines species specific chemical characteristics and previously derived relationships between hyperspectral data and foliar chemistry. Field data utilized for validation of the classification included both a stand-level survey of stem diameter, and field measurements of plot level foliar biomass. A random selection of validation pixels yielded an overall classification accuracy of 75%.
- Efficient forest resources management through GIS and remote sensing
Lakshmi, VU; Murthy, MSR; Dutt, CBS
Current Science [Curr. Sci.], vol. 75, no. 3, pp. 272-282, 10 Aug 1998
The optimal and efficient management of forest resources call for reliable technologies with a provision to store, update, retrieve and analyse data. Towards this, tools like Geographic Information System (GIS) and Remote Sensing (RS) have been used for decision making and to derive meaningful outputs for plant resources conservation and management. The potential application of GIS is illustrated through various case studies ranging from development of Forest Resources Information System at divisional level to micro-level planning in Joint Forest Management areas. The studies related to plant diversity prospecting, inputs for forest working plans, etc. have also been discussed in the paper.
- Estimation of forest leaf area index using remote sensing and GIS data for modelling net primary production
Franklin, SE; Lavigne, MB; Deuling, MJ; Wulder, MA; Hunt, ERJr
International Journal of Remote Sensing [INT J REMOTE SENS], vol. 18, no. 16, pp. 3459-3471, 10 Nov 1997
Digital Landsat TM imagery and GIS data in the Fundy Model Forest of southeastern New Brunswick were examined to determine relations to forest leaf area index within different stand structures or covertypes. The image data were stratified using GIS covertype information prior to development of leaf Area Index (LAI) from Landsat TM imagery was improved with reference to estimates of stem density. Actual stand LAI was compared to assumed maximum LAI values for several species and sites using an ecosystem process model (BIOME-BGC). Subsequent comparison of pNPP and a NPP revealed that even disturbed sites in this environment can reach close to maximum site potential.
- Mapping tropical forest fractional cover from coarse spatial resolution remote sensing imagery
Foody, GM; Lucas, RM; Curran, PJ; Honzak, M
Plant Ecology [Plant Ecol.], vol. 131, no. 2, pp. 143-154, Aug 1997
At regional to global scales the only feasible approach to mapping and monitoring forests is through the use of coarse spatial resolution remotely sensed imagery. Significant errors in mapping may arise as such imagery may be dominated by pixels of mixed land cover composition which cannot be accommodated by conventional mapping approaches. This may lead to incorrect assessments of forest extent and thereby processes such as deforestation which may propagate into studies of environmental change. A method to unmix the class composition of image pixels is presented and used to map tropical forest cover in part of the Mato Grosso, Brazil. This method is based on an artificial neural network and has advantages over other techniques used in remote sensing. Fraction images depicting the proportional class coverage in each pixel were produced and shown to correspond closely to the actual land cover. The predicted and actual forest cover were, for instance, strongly correlated (up to r = 0.85, significant at the 99% level of confidence) and the predicted extent of forest over the test site much closer to the actual extent than that derived from a conventional approach to mapping from remotely sensed imagery.
- High spectral resolution remote sensing of forest canopy lignin, nitrogen, and ecosystem processes
Martin, ME; Aber, JD
Ecological Applications [Ecol. Appl.], vol. 7, no. 2, pp. 431-443, May 1997
Remote sensing of foliar chemistry has been recognized as an important element in producing large-scale, spatially explicit estimates of forest ecosystem function. This study was designed to determine whether data from NASA's Airborne Visible/Infrared Imaging Spectrometer (AVIRIS) could be used to determine forest canopy chemistry at a spatial resolution of 20 m, and if so, to use that information to drive an ecosystem productivity model. Foliage and leaf litter were sampled on 40 plots at Blackhawk Island, Wisconsin, and Harvard Forest, Massachusetts, to determine canopy-level nitrogen and lignin concentrations. At the time of the field sampling, AVIRIS data were acquired for both study areas. Calibration equations were developed, relating nitrogen and lignin to selected first-difference spectral bands (R super(2) = 0.87 and 0.77, respectively). Calibration equations were evaluated on the basis of inter- and intrasite statistics. These equations were applied to all image pixels to make spatially explicit estimates of canopy nitrogen and lignin for both study sites. These estimates of nitrogen and lignin concentrations were then used with existing models to predict net ecosystem productivity at Harvard Forest and nitrogen mineralization rates at Blackhawk Island.
- Remote sensing of aerosols over boreal forest and lake water from AVHRR data
Soufflet, V; Tanre, D; Royer, A; O'Neill, NT
Remote Sensing of Environment [Remote Sens. Environ.], vol. 60, no. 1, pp. 22-34, Apr 1997
A complete set of advanced very high resolution radiometer (AVHRR) data and ground-based measurements of aerosol and water-vapor content are used to test an algorithm for the retrieval of aerosol properties over dense vegetation in the red and over lake water in both the red and the near-infrared AVHRR channels. With the assumptions of a weak and reasonably constant surface reflectance and an appropriate aerosol model in the radiative transfer code, the remaining variance in the satellite signal is interpreted in terms of aerosol optical thickness. From theoretical computations, it appears that the algorithm is particularly sensitive to the surface albedo and that an uncertainty of 0.01 in reflectance leads to an error of plus or minus 0.1 in the retrieved optical thickness. This theoretical estimate is confirmed by data acquired over a boreal forest region in Canada and over one of the Great Lakes (Ontario). In particular, channel 1 observations over vegetation in the forward scattering direction are well suited for retrievals because vegetation pixels appear darker owing to shadowing effects. Conversely, the forward scattering geometry over lakes introduces large errors in both channels owing to specular reflections (glint effects). Even for observations well removed from the forward scattering principal plane, lake surface reflections due to sky radiance glint have to be taken into account. Because the accuracy of the retrieval algorithm is affected by water-vapor absorption in channel 2 and by variations in lake-water turbidity in channel 1, the optimal retrieval configuration is to employ vegetation observations in channel 1. Bidirectional effects have to be considered, however, for observations in the backscatter direction.
- Remote sensing of forest fire severity and vegetation recovery
White, JD; Ryan, KC; Key, CC; Running, SW
International Journal of Wildland Fire [Int. J. Wildland Fire], vol. 6, no. 3, pp. 125-136, Sep 1996
Burned forested areas have patterns of varying burn severity as a consequence of various topographic, vegetation, and meteorological factors. These patterns are detected and mapped using satellite data. Other ecological information can be abstracted from satellite data regarding rates of recovery of vegetation foliage and variation of burn severity on different vegetation types. Middle infrared wavelengths are useful for burn severity mapping because the land cover changes associated with burning increase reflectance in this part of the electromagnetic spectrum. Simple stratification of Landsat Thematic Mapper data define varying classes of burn severity because of changes in canopy cover, biomass removal, and soil chemical composition. Reasonable maps of burn severity are produced when the class limits of burn severity reflectance are applied to the entire satellite data. Changes in satellite reflectance over multiple years reveal the dynamics of vegetation and fire severity as low burn areas have lower changes in reflectance relative to high burn areas. This results as a consequence of how much the site was altered due to the burn and how much space is available for vegetation recovery.
- Forest fire risk modelling using remote sensing and geographic information system
Jain, A; Ravan, SA; Singh, RK; Das, KK; Roy, PS
Current science. Bangalore [CURR. SCI.], vol. 70, no. 10, pp. 928-932, 1996
We present here an integrated remote sensing and Geographic Information System (GIS) approach for prioritization of the forest fire risk areas in a part of Rajaji National Park (Uttar Pradesh), India, which lies in the fragile ecosystem Siwalik Himalayas. The important factors favouring the forest fire and its spread, viz. fuel content in the forest, proximity to roads/fireline and settlement and the topography were integrated to obtain a fire risk zone map of the study area. Integration of different informations was made possible through PAMAP GIS. Almost 50% of the study area was predicted to be under very high and high risk zones, mainly concentrated in the fringes of the park. A comparison between predicted risk area and actual burnt areas during 1985-93 showed that 41.49% of the previously burnt area fell in the high and moderately high risk zones, suggesting that the approach was useful as a predictive tool.
- Remote sensing assessment of forest health in the Bohemian forests of central Europe
Entcheva, PK; Rock, BN; Lauten, GN; Cibula, WG
PROCEEDINGS OF ECO-INFORMA '96. GLOBAL NETWORKS FOR ENVIRONMENTAL INFORMATION., ENVIRONMENTAL RESEARCH INSTITUTE OF MICHIGAN (ERIM), PO BOX 134001, ANN ARBOR, MI 48113-4001 (USA), 1996, vol. 11, pp. 785-790
Current studies using Landsat TM data for assessment of forest damage in the Czech Republic and Germany have demonstrated that tree levels of forest damage (light, moderate and heavy) can be discriminated applying logit regression methods, when on the ground foresters can recognize a total of five levels of decline. Field studies using portable spectrometer and a narrow-band video camera provide evidence for recent improvement in forest health and demonstrate that monitoring of the red edge portion of the visible /near infrared region of the spectrum may provide the early warning capabilities, missing when using broad band sensor systems. Detailed measurements of initial stages of damage suggest that hyperspectral sensors, such as the Lewis HSI scheduled for launch in 1997, will provide the capabilities for detection and identification of the initial stages of forest damage.
- Assessment of forest fragmentation in southern New England using remote sensing and geographic information systems technology
Conservation Biology [CONSERV. BIOL.], vol. 9, no. 2, pp. 439-449, 1995
Spatial patterns and rates of forest fragmentation were assessed using digital remote sensing data for a region in southern New England that included 157 townships in southern New Hampshire and northeastern Massachusetts. The study area has undergone marked population increases over the last several decades. Following classification of 1973 and 1988 Landsat Multispectral Scanner data into forest and nonforest classes, data were incorporated into a geographic information system. The natural logarithms of forest area to perimeter ratios, referred to as the forest continuity index, were used to assess patterns and trends of forest fragmentation across the region. Forest continuity index values were extracted from each township for both data sets and compared with population data. Forest continuity index values were found to decrease with increasing population density until about 200 persons per square kilometer, after which the relationship stabilized. With slight population increases at low densities forest continuity index values declined sharply, implying abrupt increases in forest fragmentation. Results from the study indicated good negative correlations (r super(2) values of 0.81 and 0.77) between the Multispectral Scanner-derived forest continuity index and natural logs of township population density. Socioeconomic indicators such as affluence and commuting patterns did not appear to correlate well with forest fragmentation estimates. Decreases in forest continuity index values occurred throughout much of the study region between 1973 and 1988, suggesting that forest fragmentation is occurring over large regions within the eastern United States. It is technologically feasible to assess patterns and rates of forest fragmentation across much larger areas than analyzed in this study; such analyses would provide useful overviews enabling objective assessment of the magnitude of forest fragmentation.
- Utilization of SAR and optical remote sensing data for habitat conservation in the tropical forest of Brazil
Lawrence, William; Saatchi, Sasan; DeFries, Ruth; Dietz, James; Rice, Richard; Dietz, Lou Ann; de Araujo, MSiquiera; Alger, Keith
DIG INT GEOSCI REMOTE SENS SYMP (IGARSS), IEEE, PISCATAWAY, NJ, (USA), 1995, vol. 2, pp. 1480-1482,
This research has been undertaken with two goals: to find a suitable habitat for the expansion of forest reserve for endangered primate (the golden-headed lion tamarind); and to examine the possibility of establishing sustainable agricultural practices that would eliminate or reduce the need of further deforestation. The first effort, which attempts to classify the vegetation cover of an area using images from Landsat Multispectral Scanner, is not successful. The second data gathering effort, however, proves to be more successful as it culminates with the acquisitions of a previously archived Landsat Thematic Mapper image and new synthetic aperture radar images. These two datasets, finally, have allowed the authors to distinguish between previously confused classes, and to have insights into canopy cover, vegetation structure, and biomass for biodiversity and conservation efforts.
- Satellite remote sensing of balsam fir forest structure, growth, and cumulative defoliation
Franklin, Steven E; Luther, Joan E
Canadian Journal of Remote Sensing [CAN J REMOTE SENS], vol. 21, no. 4, pp. 400-411, 1995
Digital Landsat Thematic Mapper (TM) and SPOT High Resolution Visible (HRV) sensor data were compared in their relationship to Balsam Fir (Abies balsamea (L.) Mill.) forest stand growth parameters, stand structure measured in the field and stand parameters in the provincial forest inventory, and defoliation caused by the blackheaded budworm (Acleris variana (Fern.)) in western Newfoundland. Correlation, multiple regression, and logistical regression techniques were used to determine the relationships and interpret their value in the context of forest management in Newfoundland. Weak, but statistically significant relationships were found between multispectral reflectance and basal area (SPOT only), diameter (TM only), volume, volume increment (TM only), tree vigour, and growth efficiency. In most cases Landsat TM provided stronger relationships than SPOT data, probably because of the lower dynamic range and spectral dimensionality of the SPOT sensor. The relationships between spectral reflectance and forest parameters generally were stronger when growth measures, such as growth efficiency or volume increment, were considered compared to structural measures, such as density or diameter, probably because the growth measures can serve as summary indices of several aspects of stand structure. However, still stronger relationships were observed between spectral reflectance and forest inventory data categorized into broad structural classes. In the final test, SPOT HRV imagery were reasonably successful in detecting cumulative defoliation with approximately 70% accuracy three years after the final year of the infestation.
- Forest condition and forest damages - Contribution of remote sensing to different inventory approaches
GeoJournal, vol. 32, no. 1, pp. 47-53, 1994
The development of inventory activities in the field of forest damage assessment and monitoring during the last decade in Germany and the present state are recorded as far as remote sensing has been involved. Any forest inventory is influenced by external factors, and the resulting difficulties for an introduction of new technologies are described. The following tasks and/or methods are discussed: - global approaches to deforestation monitoring - working experience from local and regional case studies - vegetation and vegetation-damage monitoring in "urban forestry" - sampling approaches for large areas - the contribution of spectral signatures and satellite remote sensing to damage assessment.
- The development of forest damage - Control and prognosis on the basis of remote sensing data
Heiner, B; Uta, S; Rainer, S
GeoJournal, vol. 32, no. 1, pp. 39-46, 1994
When using remote sensing data to monitor the development of forest damage it is necessary not only to pay attention to spectral signatures characterizing the degree of green space and thus the information of the canopy but also to spectral signs showing the water supply of the stands. This has not been done so far though nowadays the necessary multispectral satellite data are on regular offer. For this reason it may be suggested that the existing possibilities to predict the development of forest damage on the basis of remote sensing data have not yet been fully used. In order to examine these possibilities, in the environs of Berlin, ie in the east and south of the Land of Brandenburg both the vegetation index which has been known from literature for a long time and an index for the difference in remission identifying the water bands in the short-wave infrared have been calculated on the basis of Landsat-TM data. A multi-temporal comparison shows that despite the bad conditions of the forest in 1991 on the test site located north-east of Berlin due to a relatively good water supply in that year the pine forest damage had not progressed in 1992. This allows to draw the conclusion that at least in the pine forests of the East German interior lowland the development of forest damage is essentially determined by dryness stress. In this way on the basis of 1992 Landsat-TM data there was not only an inventory made of the actual condition of the forest on the east and south-east of the Land of Brandenburg but also trends of damage progress shown. The damage classification of pine stands be completed by a prognosis as to the development of damage.