Environmental Engineering Reference
In-Depth Information
However, since resulting forest wood harvesting aggregated statistics appar-
ently underestimate real wood production, the whole reporting system was fre-
quently criticized in literature (Cutolo
2000
; Corona et al.
2004
). Forest owners
may in fact communicate clearcuts smaller than the effective ones avoiding the
cost of a formal harvesting project and simply cutting in a larger area. Just a few
clearcuts are verified in the field by local authorities.
It is evident that since updated forest wood harvesting statistics are so impor-
tant in supporting strategic forest planning actions, an accurate, cheap and feasible
monitoring system should be implemented. The integrated use of remote sensing
technologies should be considered. The results recently published in Chirici et al.
(
2011
) are here synthesized.
7.2.2 Remote Sensing for Harvesting Monitoring
Monitoring of forest disturbances by remote sensing is usually implemented
through the use of change detection techniques (e.g. Coppin et al.
2004
; Lu et al.
2004
).
Forest harvesting monitoring through satellite image analysis is reported for the
tropics (e.g. Souza et al.
2005
), USA (e.g. Schroeder et al.
2007
), Canada (e.g.
Wulder et al.
2007
), and boreal Europe (e.g. Heikkonen and Varjo
2004
).
Chirici et al. (
2011
) presented the first study specifically oriented to satel-
lite monitoring of wood harvesting in coppice forests. The authors tested man-
ual and semi-automatic methods to define a correction coefficient to be applied
to official statistics in order to obtain a more reliable data on annual harvested
extents.
The test was carried out within the framework of the project Global Monitoring
for Environment and Security (GMES) Service Element Forest Monitoring in a
test area of 34,000 km
2
located in central Italy. The area was covered by multitem-
poral SPOT5 HRG images (Fig.
7.2
).
Coppice forests were dominated by oaks managed with a rotation period of
about 20 years. Chirici et al. (
2011
) compared the visual on-screen photo-interpre-
tation of infrared false color composite SPOT5 HRG images with semi-automatic
classification systems based on both pixel-based and object-based approaches.
The results demonstrated that mapping and dating forest clearcuts is possible
and feasible if post-cut images are available within a timeframe of 1 or 2 years
since the cut (Fig.
7.3
).
The comparison between the in-field validated satellite-based map with over
9,500 clearcuts and aggregated official statistics demonstrated a significant under-
estimation by the latter (65 % of the total mapped clearcut area).
The object-oriented approach resulted in better results confirming the findings
of Descl￩e et al. (
2006
). An operative probabilistic procedure based on VHR (very
high resolution) satellite samples was proposed by the authors to correct official
harvesting statistics.
