Document Sample
                                                                                  Subproject 4.1

                                                        Scientific responsible: Vittorio Tosi
         (C.R.A. – Research unit for forest monitoring and management, Trento)

                                         Final report

Title: 4.1.4.- Potential of monitoring forest land by new satellite
sensors with high and very high spatial resolution

Operating Unit: DISAFRI – Università degli Studi della Tuscia, Viterbo

Responsible of O.U.: Piermaria Corona ( )


        Nowadays, remotely sensed (RS) imagery available for forest vegetation monitoring is
numerous. Up to recent times the primary RS reference for forestry applications has been
represented by aerial photos. However, they are generally of more complex acquisition in
comparison to the satellite images, at least in the case of a periodical monitoring of wide
areas. Hence, the use of satellite RS has grown quickly for forestry applications, especially to
support reliable multitemporal surveys, in terms of both qualitative, such as specific
composition or development stages of the forest stands, and quantitative attributes, such as
wood biomass or canopy coverage. In this perspective, specific targets of this project have
- assessment of the potential of optic satellite sensors with high and very high spatial
resolution aimed to map forest types stands, with a reference scale of 1:25000/1:10000 and a
detail compatible with the FAO forest definition (minimum mapped unit equal to 0.5 ha);
- evaluation of their technical capability to monitor forest land changes (e.g. forest fires).
Materials & methods

       In the last decades Earth Observation (EO) systems have been operationally configured
to provide information with high quality also at medium and high landscape scales. Main
operational applications in forestry are based on satellite optic images (passive) with high and
very high geometric resolution (e.g.: Landsat 7; Aster; IRS-1C; Spot4; Spot5; Quickbird;
Ikonos). This project has focused on Landsat7, Spot5, Quickbird and Ikonos imagery.
   The work has been organized to comparatively assess the operational potential of the
above-mentioned information sources, in comparison with the use of conventional aerial
photos too, setting up the more appropriate methods of image processing. The work has been
modulated upon the following steps:
   -   check of the use of Spot5 images for forest stand inventory with a detail compatible to
       the FAO forest definition;
   -   assessment of Ikonos and Quickbird imagery for burnt area mapping;
   -   set up of procedures of multi-resolution segmentation of Ikonos and Quickbird images
       for a multilevel thematic analysis of forest ecosystems attributes;
   -   set up of procedures of integration between information from Landsat7 images and data
       from forest inventory plots by non-parametric techniques like kNN, with the purpose of
       the spatialization (mapping) of the field-measured dendrometric attributes.

Achieved results

       From a scientific point of view innovative results have been above all of methodological
nature and of procedure standardization, with reference to the efficiency and to the operational
implementation of the tuned techniques and tools.
       The obtained results are described in details in the eleven publications produced and
discussed during twenty-one oral presentations at seminars and conferences. From a practical
point of view, it can be distinctively highlighted:
- the positive verification of Spot5 images to map the FAO forest definition, with a meaningful
potential in terms of efficiency (accuracy/cost ratio) for inventory aims (e.g. to support the
first phase of multiphase inventories);
- the multiresolution segmentation procedures and object-oriented classification of forest
stands have been demonstrated to be able to efficiently exploit the high informative content of
the geometric component of the very high resolution satellite imagery like Ikonos and
QuickBird; for the land use/cover maps, the experimentation carried out produced a mapping
protocol based on a mixed approach of supervised pre-classification of the segmented
polygons followed by a check and manual revision by on screen interpretation;
- the set up of procedures based on kNN (k nearest neighbours) algorithms have been
demonstrated to be significantly efficient for the spatialization (continuous mapping) of forest
inventory sample plot data, where the remotely sensed data (the reflectances in the spectral
bands) is used as auxiliary information.


       The project has fulfilled the project goals.
       The accomplished analyses and the achieved results highlight the necessity of further
in-depth elaborations, mainly taking into account the fast technological advancement of the EO
sector, able to quickly change any reference framework: e.g. the next launching of the
WorldView sensor, evolution of QuickBird, with a spatial resolution of 50 cm in the
panchromatic band and 2 m in the multispectral bands and that will detect in eight bands
(instead of the four of QuickBird).
       As concerns thematic issues, the potential of EO with high and very high resolution is
connected to the following requirements:
- the widespread of clearcuts on coppice stands; among the official statistics, forest harvesting
data are those with smaller accuracy in Italy and many other countries; in the light of this, the
monitoring of the clearcut surfaces by satellite images with very high resolution can give a
particularly effective contribution;
- the monitoring of the biodiversity through a multilevel analysis within the forest ecosystems
can find effective support by multiresolution segmentation procedures of satellite imagery with
very high resolution;
- the monitoring of atmospheric carbon fixed by the forest stands can be supported by
modelling applications driven by satellite imagery with high resolution.

Organization and financing

       The research has been carried out with the cooperation of the researchers listed in
Table 1.
        NAME           QUALIFICATION       COMPANY        TELEPHONE               MAIL
CORONA            Professor, responsible DISAFRI        - 0761-
Piermaria         of the O.R.U.           Unitus         357425
PORTOGHESI        Professor               DISAFRI       - 0761-
Luigi                                     Unitus         357406
CECCHI Piero      Technician              DISAFRI       - 0761-
                                          Unitus         357406
CHIRICI           Professor               STAT          - 0865
Gherardo                                  Unimol         478904
TRAVAGLINI        Research grant          DISTAF        - 055-
Davide                                    Unifi          30231221
LAMONACA          Research grant          DISAFRI       - 0761-
Andrea                                    Unitus         357417

Table 1 – List of people involved in the study 4.1.4.

         The financing available for this study is altogether equal to     49.700,00 €, with the
following partition:
        2001 = 18 500 euro (= 39 000 000 Italian lira)
        2003 = 19 200 euro
        2005 = 12 000 euro

Knowledge spreading and publications

         The results of the study were presented at several special meetings organized for the
project and at other national workshops. Among others we mention the congress               SISEF of
Viterbo (2003) and the conclusive seminary of the subproject held in Trento on the 8 th of May


CHIRICI G., CORONA P., MARCHETTI M., MASELLI F., BOTTAI L., 2003 - Spatial distibution modelling of
forest attributes coupling remotely sensed imagery and GIS techniques – In: Amaro A., Reed
D., Soares P. (a cura di), Modelling forest systems. CAB International, Wallingford, pp. 41-50.
CHIRICI G., CORONA P., MARCHETTI M., PORTOGHESI L., TRAVAGLINI D., 2003 - Feasibility assessment
of very high resolution satellite imagery for hierarchical classification of burnt pine stands
under a Mediterranean coastal environment - In: De Angelis P., Macuz A., Bucci G., Scarascia
Mugnozza G. (a cura di), Alberi e foreste per il nuovo millennio, 3° Congresso SISEF, Viterbo,
pp. 67-72.
CHIRICI G., CORONA P., MARCHETTI M., RICOTTA C., TRAVAGLINI D., 2002 - Esperienze di ricerca sul
monitoraggio mediante telerilevamento degli ecosistemi forestali a supporto di una gestione
sostenibile – Rivista Italiana di Telerilevamento 24: 69-79.
CHIRICI G., CORONA P., MARCHETTI M., TRAVAGLINI D., 2003 - Rilevamento campionario e
spazializzazione cartografica della necromassa legnosa in biocenosi forestali – Monti e Boschi
6: 40-45.
CHIRICI G., CORONA P., MARCHETTI M., TRAVAGLINI D., 2003 - Testing Ikonos and Landsat 7 ETM+
images for stand-level forest type mapping: a case study under Mediterranean conditions - In:
Corona P., Koehl M., Marchetti M. (eds.), Advances in forest inventory for sustainable forest
management and biodiversity monitoring, Kluwer, Dordrecht, pp. 71-85.
CORONA P., DEL FAVERO R., MARCHETTI M., 2004 - Stand-level forest type approach in Italy:
experiences from the last twenty years – In: Marchetti M. (a cura di), Monitoring and
indicators fo forest biodiversity in Europe. Forma ideas to operationality. EFI Proceedings 51,
pp. 143-152.
LAMONACA A., CORONA P., 2007 - Analisi multilivello della diversità strutturale di una faggeta con
caratteri di vetustà - Italia Forestale e Montana 3: 177-194.
TRAVAGLINI D., 2004 - Trasformazioni tra sistemi di coordinate: software disponibili, limiti e
potenzialità - Forest@ 2: 128-134.
immagini telerilevate Spot5 per inventari forestali su vasta scala – Italia Forestale e Montana
4: 301-317.

Articles in press

CORONA P., CHIRICI G. - Quantificazione, qualificazione e monitoraggio delle aree boschive
percorse da incendio. Utilità del telerilevamento – Atti DIMAF, Arrone.
Abstract - Qualification of the burnt scar perimeter mapping: remote sensing usefulness
Monitoring of forest burnt areas has several aims: to locate and estimate the extent of such
areas; to assess the damages suffered by the forest stands; to check the ability of the
ecosystem to naturally recover after the fire; to support the planning of reclamation
interventions; to assess the dynamics (pattern and speed) of the natural recovery; to check
the outcome of any eventual restoration intervention. Remote sensing is an important source
of information to support all such tasks. In the last decades, the effectiveness of remotely
sensed imagery is increasing due to the advancement of tools and techniques, and to the
lowering of the costs, in relative terms. For an effective support to post-fire management
(burnt scar perimeter mapping, damage severity assessment, post-fire vegetation monitoring),
a mapping scale of at least 1:10000-1:20000 is required: hence, the selection of remotely
sensed data is restricted to aerial imagery and to satellite imagery characterized by high (HR)
and, above all, very high (VHR) spatial resolution. In the last decade, HR and VHR passive
remote sensing has widespread, providing affordable multitemporal and multispectral pictures
of the considered phenomena, at different scales (spatial, temporal and spectral resolutions)
with reference to the monitoring needs.

LAMONACA A., PAOLO CALVANI P., GIULIARELLI D., CORONA P. - Aspetti inventariali dei rimboschimenti
di pino nero in provincia di Rieti – Rivista Italiana di Telerilevamento.
Abstract - Surveying black pine plantations in the province of Rieti (Italy)
In the last century large afforestation programs were carried out in the Appennines to recover
degraded lands, mainly by Pinus nigra plantations. Currently, many Authors have proposed
management guidelines to foster the naturalization of such woodlands. However, such an
option requires the availability of detailed surveys. The aim of this note is to propose a survey
protocol for qualifying the landscape pattern of pine reafforestations and for mapping their
growing stock on the basis of the k-Nearest Neighbor (k-NN) techniques. A case study
application by Landsat 7 ETM+ imagery is here reported concerning the Pinus nigra plantations
in the province of Rieti (Central Italy).

Oral presentations

- Sperimentazione di procedure di monitoraggio di aree forestali percorse dal fuoco. III
Congresso Nazionale SISEF, ‘Alberi e foreste per il nuovo millennio’, Viterbo, 15-17.10.2001.
- Testing Ikonos and Landsat 7 ETM+ images for stand-level forest type mapping: a case
study under Mediterranean conditions. IUFRO Conference, ‘Collecting and analyzing
information for sustainable forest management and biodiversity monitoring with special
reference to Mediterranean ecosystems’, Palermo, 4-7.12. 2001.
- Spatial distribution modelling of forest attributes coupling remotely sensed imagery and
GIS techniques. IUFRO Workshop ‘Reality, models and parameter estimation – the forestry
scenario’, Sesimbra, Portugal, 2-5.06.2002.
- Stand level forest types approach in Italy: results from the last twenty years. IUFRO
Conference, ‘Monitoring and indicators of forest biodiversity in Europe. From ideas to
operationality’, Firenze, 12-15.11.2003.
- Segmentazione e classificazione object-oriented di immagini telerilevate come
supporto/alternativa alla fotointerpretazione. Seminario di aggiornamento tecnico, ALIFOR,
Innovazioni in geomatica e telerilevamento per applicazioni in ambito forestale, Firenze,
- Trasformazioni tra sistemi di coordinate: software disponibili, limiti e potenzialità.
Seminario di aggiornamento tecnico, ALIFOR, Innovazioni in geomatica e telerilevamento per
applicazioni in ambito forestale, Firenze, 02.04.2004.
- Tecniche di segmentazione multirisoluzione di immagini telerilevate per la derivazione di
cartografie tematiche forestali. Seminario AIT, ASITA, AISF, ALIFOR, SBI, Università di Torino,
IPLA, ‘Utilizzo di dati telerilevati nella cartografia delle coperture vegetali’, Torino, 18.05.2004.
- Remote sensing for post-fire monitoring. Ministero Affari Esteri, FAO, International
Symposium, ‘Forest Fires, Experience from the Italian Cooperation - Prevention and active
fight in the Mediterranean’, Reggio Calabria, 24-25.11.2004.
- Stumenti del telerilevamento. Convegno, Osservatorio Foreste e Ambiente, Corpo Forestale
dello Stato, Regione Toscana, ‘Conoscere e difendere i boschi’, Vallombrosa, 9-10.06.2005.
- Analisi strutturale multilivello in un bosco vetusto di faggio. V° Congresso SISEF, ‘Foreste e
società: cambiementi, conflitti, sinergie’, Torino, 27-30.09.2005.
- Confronto di metodi parametrici e non parametrici per la spazializzazione della provvigione
legnosa tramite integrazione di misure a terra, dati telerilevati e informazioni ancillari. V°
Congresso SISEF, ‘Foreste e società: cambiementi, conflitti, sinergie’, Torino, 27-30.09.2005.
- Supporto informativo del telerilevamento per il monitoraggio e la valutazione funzionale dei
rimboschimenti come mezzi di lotta alla desertificazione. Congresso, Comitato Nazionale per la
lotta alla siccità e alla desertificazione, ‘Verso una integrazione delle conoscenze tradizionali e
nuove tecnologie per la lotta alla siccità e alla desertificazione’, Viterbo, 3-5.10.2005.
- Telerilevamento satellitare a supporto della gestione delle risorse forestali. Workshop,
Agenzia Spaziale Italiana, ‘Lo spazio al servizio del cittadino’, Roma, 30.03.2006.
- Point sampling as a tool for landscape change assessment and modeling. Spatial
Conference, ‘Spatial data methods for environmental and ecological processes’. Foggia - Baia
delle Zagare, 14-15.09.2006.
- Strumenti cartografici, inventariali e di monitoraggio a supporto della pianificazione
forestale. Giornata Internazionale della Montagna, ‘Quale futuro per montagna in Sicilia’,
Catania, 11-12.12.2006.
- Monitoraggio post-incendio della rinnovazione naturale tramite l’uso integrato di immagini
telerilevate ad alta risoluzione. Seminario, AIT, ASITA, CNR, AISF, SBI, ‘Telerilevamento e
spazializzazione nel monitoraggio forestale e ambientale per la difesa degli ecosistemi, Pesche,
- Metodi parametrici e non parametrici per la stima della provvigione legnosa a fini
assestamentali. Seminario, AIT, ASITA, CNR, AISF, SBI, ‘Telerilevamento e spazializzazione nel
monitoraggio forestale e ambientale per la difesa degli ecosistemi, Pesche, 24.01.2007.
- Monitoraggio della superficie forestale e rappresentazioni cartografiche. Seminario Progetto
finalizzato RiselvItalia, MiPAAF, ‘Inventario e monitoraggio delle risorse e degli ambienti
forestali’. Trento, 08.05.2007.
- Utilizzo di dati ASTER per la caratterizzazione di ambienti forestali. Simposio ENEA,
Università di Bologna, ‘I dati satellitari ASTER: studi territoriali’, Bologna, 09.05.2007.
- Testing object oriented based analysis to support sustainable management of forest
plantations. 27th EARSEL Symposium, ‘Geoinformation in Europe’, Bolzano, 04-07.06.2007.
- Design based approach to the kNN techniques in forest inventory: study cases. IUFRO
4.02, AISF, ‘2nd International Nearest Neighbors Workshop’, Firenze, 05-07.07.2007.

Shared By: