conclusion of satellite imagery
Answers
Answer:
Explanation:
This is the first survey of its kind to measure, in a systematic way, losses and gains
in forest land use between 1990 and 2005 at the global, regional, climatic domain and
ecological zone levels of aggregation. The results presented in this report indicate that
forest conversion to other land uses is most prevalent in the tropical climatic domain
and, within this domain, in South America. Other climatic domains were remarkably
stable in terms of net forest land-use change over the period 1990–2005.
The systematic survey design permitted estimates of gross forest area gains and
losses and net changes in forest area, each with an estimate of precision. The exhaustive
review-and-revision process by national-level forestry and remote sensing experts
made possible the correction of classification errors and the identification of land uses
not discernible from remotely sensed data sources alone, and provided an improved
ecological context for the monitoring of forest cover and forest land-use change
globally.
INTEGRATION OF COARSE RESOLUTION SATELLITE IMAGERY TO HELP
CLASSIFICATION
The survey benefited from the use of global coarse spatial resolution datasets to both
normalize and classify the relatively finer spatial resolution Landsat samples. Although
coarse spatial resolution satellite imagery is often unsuitable as a stand-alone data
source for detecting change, several studies have shown the effectiveness of using such
data for the purpose of selecting training data for land-cover classifications at finer
spatial resolutions. For example, Hansen et al. (2008) showed the utility of using coarse
spatial resolution data from the MODIS VCF product to delineate potential training
sites for a forest/non-forest classification in Central Africa. Similar methods have also
been applied successfully in the Brazilian Legal Amazon (Broich et al., 2009), Indonesia
(Broich et al., 2011), and the boreal region of the Russian Federation (Potapov et al.,
2008; Potapov, Turubanova and Hansen, 2011).
Importance of visual review and revision of classification
Visual control and correction was an important part of the land-cover and land-use
classification processes and had a large impact on the final results. A comparison of
the initial results from the automated land-cover classification and final reviewed-andrevised results for the tropics indicated that about 20 percent of the polygon labels were
revised by national experts (Raši et al., 2011). Similar results were obtained for sites in
the boreal, temperate and subtropical domains (Lindquist et al., submitted). The visual
refinement process also had a notable effect on estimates of forest area and forest area
change: for Southeast Asia, for example, the net rate of change in tree cover (loss) from
1990–2000 was assessed at 0.9 percent before and 1.6 percent after visual control (Raši
et al., 2011).
THE UTILITY OF LANDSAT FOR GLOBAL MONITORING
Land-cover classification and change detection methods that leverage available data
from the current generation of Landsat sensors is critical for maintaining a record of
land-cover changes until the new generation of sensors comes online. The Landsat
programme has the longest continuous time-series of similar remotely sensed Earth
observations and is a critical component in the analysis of change in land cover and
land use since the 1970s. Landsat 7, the latest sensor, was launched in 1999 but suffered
28 Global forest land-use change 1990–2005
a mechanical failure in May 2003 that created no-data gaps in the across-track scan line
covering 23 percent of each image (Williams, Goward and Arvidson, 2006). Sampling
methods, such as those described in this report, are a suitable use of the currently
available Landsat image acquisitions and should be used to leverage the large amounts
of information freely available in the Landsat archive (Woodcock et al., 2008).
ESTABLISHMENT OF GLOBAL NETWORKS
The project established two very important global networks. One was the global
survey grid, which will be updated with data from 2010 as part of the next FRA (to
be released in 2015). The second and perhaps more important network comprises
the many national experts who participated in the survey and who remain important
points of contact and sources of forest remote sensing and land-use expertise in
individual countries.
Answer:
hfdfkkkurdx nlkgfygx-mzgndjlsngxjhsushjshkdmhshkshksbmz