VEGETATION 2000



[pic]

VEGETATION 2000

Belgirate, April 3-6, 2000

PRELIMINARY AGENDA

(03/23/2000)

TIME TABLE

Monday April 3rd

|14:00 |Opening and Welcome |JRC/SAI Representative |

|14:15 |KEYNOTE ADDRESS |B Moore Chair SC-IGBP |

|15:00 | The VEGETATION Programme, status and future. |D Faivre, A Ghazi, Cochairmen of the VEGETATION Steering |

| | |Committee |

|15:45 |VEGETATION standard pre-processing functions |X Passot (Integrated Project Team) |

|16:15 |Product quality performance |P Henry (CNES) |

|16:45 |Production entity |D Van Speybroek (CTIV) |

|17:15 |Distribution entity |S Moreau (SPOT Image) |

PLENARY SESSIONS

Tuesday April 4th

09:00 Session 1

UTILIZATION OF VEGETATION TO EXTRACT EFFECTIVE SURFACE PARAMETERS 9

Jiaguo Qi, 9

Estimation of land surface albedo and vegetation biophysical properties using SPOT-4 VGT and semi-empirical BRDF models. 10

M. J. Barnsley, T. L. Quaife, P. D. Hobson and J. Shaw, 10

VEGETATION/SPOT for Northern Applications: Assessment of Utility and Examples of Products 11

Jing Chen and Josef Cihlar 11

Validation of neural network techniques to estimate canopy biophysical variables from VEGETATION data 12

M.Weiss(1) , F.Baret(1), M.Leroy(2), O.Hautecœur(2), L.Prévot(1), and N.Bruguier(1) 12

11:00 Session 2

Estimation of Net Primary and Net Ecosystem Productivity of European terrestrial ecosystems by means of the C-Fix model and NOAA/AVHRR data. 13

Frank Veroustraete and Hendrik Sabbe 13

Monitoring North American Grasslands Dynamics with VEGETATION 14

David J. Meyer *, 14

Multitemporal analysis of the VEGETATION data for landcover assessment and monitoring in Indochina. 15

Chandra Giri & Surendra Shrestha 15

The Suitability of VEGETATION for Mediterranean Land Degradation and Desertification Monitoring 16

W. Mehl, P. Strobl, S. Sommer, H. Bohbot 16

14:00 Session 3

Monitoring of forest ecosystems at regional scale using VEGETATION daily-data : First results on the Landes maritime pine forest (SW France) 17

Jean-Pierre Lagouarde 1,Dominique Guyon 1, Benoît Duchemin 2 17

The potential contribution of SPOT 4/VEGETATION data for mapping Siberian forest cover at the continental scale 18

S. Bartalev (1), F. Achard (1), D. Erchov (2) and V. Gond (1 18

Fire Scar Detection in the Canadian Boreal Forest 19

Plummer, S.E., Gerard, F.F. and Wyatt, B.K. 19

Monitoring Boreal Forest Resources in Northern Europe from the VEGETATION instrument 20

Bernard Pinty(1), 20

16:00 Session 4

SPATEM: The analysis of annual sequences of VEGETATION data at the landscape scale. 21

Agustin Lobo and Nicolau Pineda 21

VEGETATION/SPOT4 applications for macro-regional landscape mapping 22

Lioubimtseva E. (1), 22

Fire patches in natural vegetation in southern Africa 23

Swinnen E. *, Verwimp R. **, Gulinck H. *** 23

Application of SPOT 4-VEGETATION data for mapping the forest-cover of Madagascar 24

Mayaux Philippe, Gond Valéry and Bartholomé Etienne 24

Wednesday April 5th

09:00 Session 5

STEM-VGT : Satellite measurements and terrestrial ecosystem modelling using VEGETATION instrument 25

G. DEDIEU (LERTS/ CESBIO Toulouse France) 25

Intermediate Scale Approach for Estimating Vegetation Canopy Leaf Area Index using SPOT4/VGT Spectral Bands. 26

F. Cipriani, E. Cubero-Castan 26

A new vegetation map of Central Africa 27

Herman Eerens, Bart Deronde & Jan Van Rensbergen 27

Sub-pixel mapping of Sahelian wetlands using multi-temporal SPOT-VEGETATION images 28

Jan Verhoeye, Robert De Wulf 28

Estimation of surface variables at the sub-pixel level for use as input to climate and hydrological models 29

Jean-Pierre Fortin*,Monique Bernier , Ali El Battay, Yves Gauthier and Richard Turcotte 29

11:00 Session 6

Integration of VEGETATION and HRVIR data into yield estimation approach. 30

André HUSSON 30

Interest of MIR data from VEGETATION for the monitoring of climatic phenomena impact on crops, a case study 31

Thierry Fourty 31

VEGETATION contribution to the desert locust habitat monitoring 32

Cherlet Michael* , Mathoux Pierre**, Bartholomé Etienne*** and Defourny Pierre** 32

Mediterranean habitats: a multi-variate analysis of VEGETATION data. 33

Agustín Lobo1, Jordi Carreras2 and Josep-Maria Ninot2 33

14:00 Session 7

SENSITIVITY ANALYSIS OF COMPOSITING STRATEGIES: MODELLING AND EXPERIMENTAL INVESTIGATIONS 34

de Wasseige Carlos*, Lissens Gil**, Vancutsem Christelle*, Veroustraete Frank** and Defourny Pierre* 34

Modeling directional reflectance in rugged terrain using VEGETATION products 35

Lihong Su, Xiaowen Li, Jindi Wang 35

Development of a spectral index optimized for the VEGETATION Instrument 36

Michel M. Verstraete, Nadine Gobron and Bernard Pinty 36

MC-FUME: A new method for compositing individual reflective channels 37

GiL Lissens, Els Brems and Frank Veroustraete 37

SPACE-VEGETATION SOFTWARE : A software for pre-processing VEGETATION L-Band Data 38

César CARMONA-MORENO 38

16:00 Session 8

Detection of Clouds and Cloud-Shadows for VEGETATION images 39

Pieter KEMPENEERS, Gilbert LISSENS, Freddy FIERENS, Jan VAN RENSBERGEN 39

NEW ALGORITHMIC CONCEPT FOR ATMOSPHERIC AND DIRECTIONAL CORRECTION OF THE SURFACE REFLECTANCES 40

P. MAISONGRANDE 40

Round Table on New Standard Products

Thursday April 6th

9:00 Session 9

European Forest Mapping using VEGETATION data 41

Hervé Jeanjean, 41

Mapping and monitoring small ponds in dryland with the VEGETATION instrument – application to West Africa 42

V. Gond*, E. Bartholomé*, F. Ouattara°, A. Nonguierma+ 42

Detection and mapping of burnt areas and active fires in tropical woodland ecosystems with the VEGETATION sensor: the SMOKO-FRACTAL case study over Northern Australia 43

D. Stroppiana1, M. Maggi1, J-M. Pereira2, D. Graetz3, J-M. Grégoire1, J. 43

DISTURBED ECOSYSTEMS DYNAMICS IN THE ARAL SEA REGION BY REMOTE SENSING AND GIS METHODS. 44

R. Ressl, A. Ptichnikov, G. Kapustin, P. Reimov, D. Forstman. 44

11:00 Session 10

Crop Growth Monitoring with Coupling of AVHRR and VEGETATION 45

Wu Bingfeng 45

Use of medium-resolution imagery in the Belgian Crop Growth Monitoring System (B-CGMS) 46

K. Wouters*, H. Eerens*, D. Dehem**, B. Tychon**, D. Buffet*** & B. Oger*** 46

Combined use of VEGETATION and RADARSAT data for updating snowpack cover and water equivalent in the HYDROTEL hydrological forecasting model 47

Monique Bernier*, Jean-Pierre Fortin, Yves Gauthier, Richard Turcotte and Ali El Battay 47

Antarctic snow characteristics from POLDER and VEGETATION data 48

Michel FILY, Olivier MANSE, Jean-Pierre BENOIST 48

14:00 Session 11

Applications of VEGETATION data to resource management in arid and semi-arid rangelands 49

Bégué (CIRAD),G. Chehbouni (IRD), R. Escadafal (IRD) 49

VEGETATION potentialities in food early warning systems in the Sahelian region 50

TYCHON Bernard , OZER Pierre and TOURE Souleymane 50

Incorporating the use of VEGETATION data in FAO’s programmes 51

F.L. Snijders 51

The contribution of VEGETATION/ SPOT 4 products to Remote Sensing Applications for Food Security, Early Warning and Environmental Monitoring in the IGAD sub-region. 52

Guy PIERRE SCOT 52

The Millennium Land Cover Assessment initiative 53

S. Belward 53

16:00 Conclusion

Jean Paul Malingreau (Chairman of the VEGETATION International Users Committee)

Rudolf WINTER (Director, Space Application Institute, JRC)

POSTER PAPERS

Monitoring natural disasters and “hot spots” of land-cover change with SPOT VEGETATION data to assess regions at risk and vulnerability 55

Prof. E. F. Lambin, Dr. I. Reginster & F. Lupo Sartor 55

Improved atmospheric corrections and data compositing methods for surface reflectance retrieval 56

Ph. Maisongrande (1), B. Duchemin (1), M. Leroy (1) (P.I.), G. Dedieu (1), J.L. Roujean (2), B. Berthelot (1), Ch. Dubegny (1), R. Lacaze (2) 56

VALIDATION OF BIOPHYSICAL PRODUCTS DERIVED FROM LARGE SWATH SENSORS FOR GLOBAL BIOSPHERE MONITORING 57

F. Baret(1) 57

Improving access to VEGETATION data: some results of on-going experiments 58

E. Bartholomé*, V. Gond*, S. Morimondi* 58

INTERCOMPARISON OF DEKADAL VEGETATION INDEX FROM NOAA/AVHRR AND SPOT4/VEGETATION OVER THE IGAD REGION 59

T. Bennouna and P. Bicheron 59

Vegetation Action & Demonstration Plan for desertification monitoring in China 60

Christian CREPEAU SCOT 60

Vegetation Action & Demonstration Plan for dry grassland monitoring in Senegal 61

Christian CREPEAU SCOT 61

Classifying land cover types with VEGETATION data in dryland: A case study in Burkina Faso 62

V. Gond, E. Bartholomé 62

UTILISATION DES DONNEES VEGETATION POUR LE SUIVI DE LA CAMPAGNE AGROPASTORALE SUR LA ZONE CILSS 63

A ROYER 63

Generating fine spatial resolution VEGETATION derived imagery using SAR 64

Conrad M. Bielski, François Cavayas and Langis Gagnon 64

Sub-pixel characterization of land cover at the global scale using SPOT-VEGETATION imagery () 65

Else Swinnen (*), Frank Canters(**) & Herman Eerens (*) 65

L'Établissement de Nomenclatures “Végétation” à partir d'Images SPOT 66

D. Blamont, M. Raffy 66

UTILISATION DE SPOT4-VEGETATION POUR L’ETUDE DU CHANGEMENT D’ECHELLE. 67

M. Raffy, D. Blamont 67

VEGETATION data for monitoring woody vegetation in landscape frameworks 68

Hubert Gulinck and Tim Wagendorp 68

VEGETATION data for regional forest cover mapping of Southeast Asia 69

H-J Stibig, R. Beuchle, V. Gond 69

An evaluation of SPOT-VEGETATION, for land cover mapping and the evaluation of forest resources, I: Mato Grosso, Brasil. 70

Jones, S. D. Eva, H. D., 70

BRDF correction in SPOT 4/VEGETATION ten-days composite imagery for mapping of boreal forest 71

D. Erchov (1), S. Bartalev (2), M. Deshayes (3), J. R. Dymond (4) 71

Detecting active fires with the VEGETATION instrument 72

V. Gond*, M. Maggi*, P. Henry°, J.-M. Grégoire*, E. Bartholomé* 72

Drawbacks and advantages of the VEGETATION and AVHRR instruments for burnt area detection in Northern Australia. 73

D. Stroppiana1, M. Maggi1, D. Graetz2, S. Campbell2, I. Balzer2, J-M. Grégoire1 and J.M.C Pereira3. 73

Using SPOT 4 HRVIR and "VEGETATION " sensors to assess impact of tropical forest fires in Roraima 74

T. PHULPIN, F. LAVENU, M. F. BELLAN, B. MOUGENOT and F. BLASCO 74

A Large Forest Fire in the Mediterranean region as seen by VEGETATION 75

Agustin Lobo and Nicolau Pineda 75

ATMOSPHERIC MESOCYCLONES OVER POLAR SEAS AND THEIR INFLUENCE ON ECOLOGICAL REGIME FORMATION 76

Lagun V.E., Lutsenko E.I. 76

INVESTIGATION OF RUSSIAN ARCTIC ECOSYSTEMS VARIATION CAUSED BY ANTROPOGENIC ACTIVITY AND CLIMATE CHANGE 77

Ivanov V.V., Lagun V.E. 77

Mapping Biological Diveristy in Boreal Forest From Space Using Ecological Models 78

Anthony J. Warren, Michael J. Collins 78

EVALUATION OF VEGETATION CLOUD MASKS FOR CLIMATOLOGY STUDIES AND DESIGN OF SATELLITE SYSTEMS 79

J. Hamon, L. Harang, A. Rodot 79

An ICT-based course in Earth Observation with emphasis on VGT-data 80

Rombout Verwimp1c, Ann Willekens1, Jos Van Orshoven1 and Jan Elen2 80

PLENARY SESSIONS

UTILIZATION OF VEGETATION TO EXTRACT EFFECTIVE SURFACE PARAMETERS

Jiaguo Qi,

Michigan State University

Gerard Dedieu, Yann Kerr, CESBIO, Toulouse, France,

Ghani Chehbouni, IMADES/ORSTOM, Mexico

Initial effort of this investigation was to characterize biophysical properties in the arid and semiarid southwest United States with VEGETATION data. Soon after the imagery became available, its applications expanded to include areas of Great Lakes and tropical Amazon regions where deforestation is drawing attention worldwide. Although continued effort is in process, some results from these VEGETATION images have been achieved and showed a great potential of the VEGETATION imagery in characterizing terrestrial surfaces. This report contains past and ongoing field activities and some results from VEGETATION imagery analysis. In particular, we will report in the following three areas: 1) practical techniques for correcting bidirectional effect found in the VEGETATION imagery with a focus on field activities and modeling effort, 2) feasible and practical alternatives to circumvent atmospheric effects, and 3) operational use of the data to retrieve biophysical variables such as fractional green and senescent vegetation cover and green leaf area index at variable spatial scales. Research effort is continuing and some peer-reviewed articles have been or are being published.

Estimation of land surface albedo and vegetation biophysical properties using SPOT-4 VGT and semi-empirical BRDF models.

M. J. Barnsley, T. L. Quaife, P. D. Hobson and J. Shaw,

Earth Observation & Environmental Monitoring Group,

Department of Geography,

University of Wales Swansea, U.K.

P. Lewis and M. I. Disney,

Remote Sensing Unit,

Department of Geography,

University College London, U.K.

The VGT instrument facilitates analysis of the land-surface angular signature by virtue of its wide swath width ( multiple over-passes of a location provide image data at varying sun-sensor geometries.

In recent years considerable effort has been directed towards devising methods to assess the information content of such data, most notably in the formulation of models of the Bidirectional Reflectance Distribution Function (BRDF). The three commonly cited potential products of BRDF model inversion are 1) the radiometric normalisation of off-nadir pixels, 2) the improved estimation of narrow- and broadband albedo and 3) the estimation of land-surface biophysical properties such as Leaf Area Index (LAI). Semi-empirical linear models of the BRDF — such as the AMBRALS suite proposed for use with the MODIS sensor — are rapidly invertible and have the potential to provide the aforementioned outputs.

Data have been collected from the VGT sensor for the period 17th of May to1st of October 1999, covering Western Europe. This constitutes a total of 478 scenes (all P level products). The data were atmospherically corrected in-house using the Simplified Method for Atmospheric Correction (SMAC) and cloud masked using the reflectance information in the blue channel.

Three areas have been selected for detailed study:

1) East Anglia, UK — a MODLAND validation site with which some of the authors are involved, and for which field data is available. The area is flat with large arable farming areas.

2) Bordeaux, France — Focusing on the Les Landes area, and

3) SE Spain — a semi-arid area with low cloud-contamination.

The AMBRALS BRDF model was inverted using VGT data for the above scenes. The potential for angular normalisation and estimation of spectral albedo is clearly demonstrated. Investigation of the potential for biophysical parameter and broad band albedo estimation is ongoing -- the main focus of which is fieldwork in East Anglia, including investigations with plant growth models and the temporal trajectories of the BRDF model parameters.

VEGETATION/SPOT for Northern Applications: Assessment of Utility and Examples of Products

Jing Chen and Josef Cihlar

Canada Centre for Remote Sensing

588 Booth Street

Ottawa, Ontario K1A 0Y7, Canada

Tel: (613)947-1266 Fax: (613)947-1406 Email: jing.chen@ccrs.nrcan.gc.ca

and

Larry Band, University of Toronto, Canada

Raymond Desjardins, Agriculture Canada

Samual Goward, University of Maryland, USA

Zhanqing Li, Canada Centre for Remote Sensing

Alain Royer, University of Sherbrooke, Canada

Robert Fraser, Intermap Technologies Ltd.

Rasim Latifovic, Canada Centre for Remote Sensing

Goran Pavlic, Intermap Technologies Ltd.

10-day Canada-wide synthesis images of VEGETATION acquired in the growing season (April 1 to November 30) in 1998 have been assessed for northern applications. The images received from the VITO centre after atmospheric corrections have been normalized to a common illumination and observation geometry (45( solar zenith angle and nadir view) after an angular correction procedure. Radiometry and atmospheric correction have been analysed using a dense dark vegetation inversion. Pixels contaminated by subpixel clouds were detected and replaced through temporal interpolation. The corrected images have been used for various applications including landcover classification, leaf area index (LAI) retrieval, fire scar area and age estimation, and net primary productivity (NPP) modelling. The utility of the VEGETATION sensor is assessed in comparison with AVHRR and Landsat TM sensors, with emphasis on the usefulness of the shortwave infrared (SWIR) channel. It is found that SWIR reflectance is very useful for improving landcover classification. SWIR information can also be used to suppress the background (understory, litter and moss) effects on forest LAI retrieval. These improvements in landcover and LAI mapping using VEGETATION images are significant in pixel-based NPP modelling. Mapping of new fire scar areas is also improved with the use of the SWIR band. The ratio of SWIR to near-infrared (NIR) reflectance was found to be highly correlated to fire scar age up to 50 years. This ability of mapping fire scar age is critical in estimating the spatial distribution of carbon sources and sinks in the northern ecosystems as the age determines the amounts of regrowth and heterotrophic respiration.

Validation of neural network techniques to estimate canopy biophysical variables from VEGETATION data

M.Weiss(1) , F.Baret(1), M.Leroy(2), O.Hautecœur(2), L.Prévot(1), and N.Bruguier(1)

(1) INRA Bioclimatologie, Domaine Saint-Paul, 84914 Avignon Cédex 9, France

(2) CESBIO, 18 avenue E.Belin, BP 2801, 31041 Toulouse Cédex 4, France

The objective of this study is to develop a global algorithm to monitor the vegetation, applicable to cultivated as well as natural vegetation areas. The monitoring is performed through the estimation of vegetation biophysical variables from 26-day VEGETATION data (fraction cover Fc, leaf area index LAI and fraction of absorbed photosynthetically active radiation fAPAR). Those variables are thus closely linked to the radiative transfer within the canopy and pertinent with regards to possible applications such as canopy primary production modeling or prediction of flux transfer of mass and at the soil-vegetation-atmosphere interface.

The learning phase of the neural networks is achieved by using a synthetic catalog of VEGETATION BRDF. The latter is built thanks to well-known radiative transfer models and a wide range of model parameters for different dates and latitudes. In this study, we do not take into account atmospheric effects and work only with top of canopy reflectance data. As the number of VEGETATION reflectance data during 26 days depends mainly from the latitude and cloud occurrence, it is necessary to pre-process these data to get a constant number of inputs required by neural networks. This is achieved by inverting a linear BRDF model to estimate the nadir and hemispherical reflectances in the 4 VEGETATION wavebands. Three neural networks are then calibrated using these inputs to estimate Fc, LAI, and fAPAR. The optimal architecture is found to be one layer with four sigmoid neurons and one output layer with one linear neuron.

A first validation is performed using a synthetic BRDF catalog of homogeneous and mixed pixels. Results show good performances on Fc and fAPAR. The LAI estimation is less satisfactory for dense canopies due to the saturation of the canopy reflectance. Moreover, LAI estimation is sensitive to the pixel heterogeneity.

A second validation is then performed on experimental data sets provided by the ReSeDA (Remote Sensing Data Assimilation, 1997). The ReSeDA site is a 4km*5km agricultural area (mainly wheat, sunflower, alfalfa and maize) near Avignon (France). Ground measurements (LAI, Fc) were performed during the whole crop cycles. Reflectance data were acquired with the airborne POLDER sensor at 15 dates during the year. The neural network technique is first modified to be consistent with POLDER measurements and then applied to retrieve biophysical variables. The comparison between estimated variables and in situ measurements is quite consistent with the results obtained with synthetic data

Estimation of Net Primary and Net Ecosystem Productivity of European terrestrial ecosystems by means of the C-Fix model and NOAA/AVHRR data.

Frank Veroustraete and Hendrik Sabbe

Flemish Institute for Technological Research (Vito)

Centre for Remote Sensing and Atmospheric Processes (TAP)

In recent years, a suite of primary productivity models has been developed, to address issues related to food security and biotic responses to climatic warming. Traditional approaches to assess primary productivity range from empirical climate correlation models to mechanistic ecophysiological models. These approaches all operate on point measurements that are extrapolated in space and time. Since landscapes are quite heterogeneous, the spatial scaling (upscaling) of point measurements is problematic, relative to the sampling density. Diagnostic or inverse (top-down) approaches to primary production like C-Fix have received a great deal of attention in recent years, owing to their avoidance of spatial extrapolation or land-cover classification, and the unique use made of remotely sensed observational data. The family of models to which C-Fix belongs to are genuinely called PEM’s (Production Efficiency Models). This means that C-Fix inherently is spatially explicit, relying on observations that are specific for a given time and location, and hence the model is driven by satellite observations.

C-Fix is a tool to create geo-referenced NPP (Net Primary Productivity) data layers for use in durable development planning and decision-making processes (Kyoto protocol). State-of-the-art algorithms describing carbon uptake and release mechanisms of carbon dioxide from vegetation cover in relation to meteorological conditions and satellite based quantification of the fraction of absorbed photosynthetically active radiation (fAPAR), allow for the build-up of yearly, geo-referenced datasets of NPP and NEP. Specific regions of interest (ROI’s) can be specified for use in GIS and land management planning environments.

In the work presented here, the modelling concepts of C-Fix will be discussed. The NOAA/AVHRR pre-processing steps to derive a pan- European fAPAR data set will be discussed. C-Fix NEP pixel simulations will be compared with data obtained from the EUROFLUX carbon exchange measurement sites. Hence the plausibility of the approach to determine NEP will be demonstrated. Finally a pan-European carbon budget will be presented and discussed within the framework of results obtained during the EUROFLUX project.

Keywords : NOAA/AVHRR, Net Primary Productivity, Net Ecosystem Productivity, C-Fix, Carbon exchange, fAPAR.

Acknowledgements: We acknowledge the financial support granted by the EU’s 4th Framework Programme under contract ENV4-CT97-0577 and the Belgian OSTC contract CG/DD/05F.

Monitoring North American Grasslands Dynamics with VEGETATION

David J. Meyer *,

Bradley K. Reed *, Bruce K Wylie *, Larry L. Tieszen +, Cullen R. Robbins *, Allison L. Scherff *

*Raytheon Systems Company, EROS Data Center, Sioux Falls, SD 57198, USA

+ U.S. Geological Survey, EROS Data Center, Sioux Falls, SD 57198, USA

This report culminates a multi-year study of the utility of the Systeme pour l’Observation de la Terre (SPOT) Haute Resolution Visible Infra Rouge (HRVIR) and Vegetation instruments for multi-scale monitoring vegetation dynamics in the Great Plains of North America. The proposed study included (1) a simulation of the Vegetation viewing geometry, including pixel size variability, (2) the use of Vegetation for monitoring biophysical parameters over grasslands, (3) the use of Vegetation for monitoring seasonal dynamics within the grasslands, and (4) the utility of having spectrally matched, simultaneously acquired multi-resolution images to study spatial scaling processes within the Great Plains. The pre-launch study focused on the simulations, refinement of field validation techniques and development of seasonal and monitoring scaling strategies for the post-launch phase; the pre-launch work is described in a previous report.

The post-launch study proceeded with biophysical, seasonal dynamics and multi-scale mapping studies using the HRVIR and Vegetation data sets. The biophysical component includes the mapping of surface relationship developed between reflectance, leaf area index (LAI), fraction absorbed photosynthetically active radiation (fAPAR) and green biomass to the HRVIR scale, then onto the Vegetation scale. The surface scaling methodology involves developing relationships between reflectance and biophysics at a “quadrat” scale (~0.5 m2), using extended reflectance “grids” large enough to be seen by the HRVIR to distribute the relationships across areas representative of HRVIR pixels, estimating the biophysical parameters at the grids scale using geostatistical techniques, then correlating the geostatistical estimates with HRVIR pixels. The scaling from HRVIR to Vegetation is done on daily syntheses acquired simultaneously with the HRVIR, the multi-date syntheses are used to distribute the measurements over time. The seasonal dynamics component of the study focused exclusively on the 10-day synetheses, using seasonal metrics develop for the Advanced Very High Resolution Radiometer (AVHRR). The results of the dynamics component of the study with Vegetation are compared to results retrieved from AVHRR data.

Multitemporal analysis of the VEGETATION data for landcover assessment and monitoring in Indochina.

Chandra Giri & Surendra Shrestha

UNEP Environment Assessment Programme for Asia and the Pacific

Asian Institute of Technology

P.O. Box 4, Klongluang, Pathumthani 12120, Thailand

Email: cpgiri@ait.ac.th & surendra@ait.ac.th

Josef Aschbacher

Directorate General Joint Research Centre (DG/JRC)

pace Applications Institute (SAI)

I-21020 ISPRA (Va) Italy

Email: josef.aschbacher@jrc.it

Land use/land cover changes are occurring at an unprecedented rate and scale in Indochina. Accurate and reliable data, however, have not been available in the past. The current study aims at improving this situation. SPOT VEGETATION and NOAA AVHRR data were used to assess the usefulness of the data for accurate delineation and demarcation of major land cover types in the region. Land cover maps of 1985/86 and 1992/93 were prepared using NOAA AVHRR and a land cover map of 2000 was prepared using VEGETATION data.

The paper presents a synopsis of this exercise focusing on the usefulness of VEGETATION data and its comparative advantage over NOAA AVHRR data. The ultimate purpose is to integrate the use of VEGETATION data into a regular assessment and monitoring operation of land cover types in Asia.

The Suitability of VEGETATION for Mediterranean Land Degradation and Desertification Monitoring

W. Mehl, P. Strobl, S. Sommer, H. Bohbot

(JRC Ispra),

R. Escadafal (IRD), J. Hill (Univ. Trier)

After launch of the VEGETATION instrument the operational parameters of the instrument and their applicability and validity for the purpose of operational monitoring of short term events as well as of pluriennal trends and indicators for land degradation are assessed. Investigated topics include:

Assessment of the product chain:

• accuracy of spatial positioning

• a strategy for identification of radiometrically stable areas, and their use for verifying the reliability of calibration and operational atmospheric correction

Assessment of alternative preprocessing steps (follow-up on prelaunch topics):

• comparison of strategies for 10 days composites

• impact of DEM-derived illumination correction on interpretability

Applications linked to high time resolution

• observation of aftereffects of rainfall on soil and vegetation in arid ecosystems

• mapping of aeolian sand transport events

Extraction of indicators relevant for long term environmental monitoring:

• seasonal evolution of sparse vegetation cover - comparison between NDVI and spectral mixture analysis methods

• in particular, assessment of state of vegetation degradation in semiarid areas based on the relationship between annual herbaceous species and perennial shrubs.

• differentiation between movable and fixed surface materials (continuation of prelaunch assessment based on simulated data)

Monitoring of forest ecosystems at regional scale using VEGETATION daily-data : First results on the Landes maritime pine forest (SW France)

Jean-Pierre Lagouarde 1,Dominique Guyon 1, Benoît Duchemin 2

(1) INRA, Unité de Recherche en Bioclimatologie, BP81, F-33883 Villenave d’Ornon Cedex

(2) CESBIO, 18 Av. Ed.Belin, F-31055 Toulouse Cedex

Previous studies carried out within the framework of Vegetation Preparatory Program on coniferous and deciduous forests using VEGETATION data simulated from AVHRR/NOAA and Landsat TM time-series. They demonstrated the potential of VEGETATION daily-data for monitoring large-scale spatial heterogeneity and temporal changes of biophysical variables which determine surface-vegetation-atmosphere transfer, net primary production (NPP), forest growth, forest yield and other environmental processes: phenological cycle duration, albedo, fraction of photosynthetic radiation absorbed (fAPAR) by forest canopy, trees-cover fraction…

We present here applications performed with actual VEGETATION daily-data for estimating these variables. The study is based on a VEGETATION daily-data set acquired during one cycle of vegetation (31 March to 7 November 1998) on the Landes maritime pine forest. It is complemented by ground measurements of forest biophysical variables and land-use geographical data base, on the test site of NEZER which covers about 3000 hectares.

The directional variations of reflectance in visible (VIS), near infrared (NIR) and middle infrared (SWIR) are analysed. A semi-empirical model of BRDF (Rahman model) already tested in the ‘pre launch’ phase of the preparatory program is fitted. Applications for the normalisation of reflectances, the estimation of albedo and fAPAR are presented and the consequences of the sampling of the BRDF during only one year discussed.

The spatial variations of reflectance and their sensitivity to forest structure parameters are also studied. VEGETATION reflectances are compared to tree-cover fraction estimated at 1km² scale from ground measurements. The practical interest of vegetation indices including the SWIR reflectance to reduce the seasonal sensitivity to phenological properties of undergrowth is analysed.

The potential contribution of SPOT 4/VEGETATION data for mapping Siberian forest cover at the continental scale

S. Bartalev (1), F. Achard (1), D. Erchov (2) and V. Gond (1

(1) Joint Research Centre of the European Commission

I-21020 Ispra (VA), Italy TP 440

( +39-0332-786396, fax +39-0332-789073, E-mail: sergey.bartalev@jrc.it

(2) International Forest Institute

117418, 69 Novocheriomushkinskaya str., Moscow, Russia

( +7-095-332 68 77, fax +7-095-332 29 17, E-mail: erchov@ifi.rssi.ru

The most recent national Russian vegetation map (i.e. covering the full territory of Russia) was published in 1990 at the scale of 1:2,500,000 (Isaev et al., 90). This map was produced from the compilation of more detailed information (inventories or maps) from various sources at different dates and with heterogeneous accuracy. For example for the Siberian North-eastern regions, the map has been derived from the spatial aggregation of visual interpretations made from aerial surveys taken in the 1950’s. Many changes in the forest cover have occurred since this period, as a consequence of natural or human-made fires, clear-cutting, insect damages and following regrowth/regeneration processes. In spite that these forests have not a large economical value, their role on the biosphere, including in the carbon cycle, is thought to be rather important.

New opportunities of getting a consistent and up-to-date forest map of Siberia are provided by the SPOT-4/VEGETATION data due to their spectral, repetitiveness and geometric characteristics well related to the issues to be solved.

A feasability study for forest mapping at the Siberian scale has been carried out using a long time series of S-10 products during the vegetative growing season: from beginning of March 1999 untill end of November 1999 covering the boreal zone of Siberia (from 42 N to 75 N and 5 E to 180E). A preliminary visual comparison between the existing Russian vegetation map and Summer mosaics of S-10 products images has already shown significant discrepancies in the map due either to initial map inaccuracies or to changes which occurred from the date of the map.

For this boreal forest mapping study at continental scale, the separability of the main forest types is investigated. These main forest types are identified by dominant tree species, such as larch (Larix spp.), spruce (Picea sibirica) and scots pine (Pinus sylvestris), … A set of locations representative of these main forest classes has been selected over the Siberian region from a few georeferenced forest maps derived from SPOT – HRV imagery. The spectral (temporal signatures) and ancillary (from inventory databases) attributes over these locations are identified and compared to each other. The separability of the classes is assessed from this comparison. The feasibility of boreal forest mapping at continental scale is further discussed.

Fire Scar Detection in the Canadian Boreal Forest

Plummer, S.E., Gerard, F.F. and Wyatt, B.K.

Centre for Ecology and Hydrology, Monks Wood, Abbots Ripton,

Cambs, PE17 2LS, UK

Tel: +44 1487 772475, Fax: +44 1487 773467, Email: sp@wpo.nerc.ac.uk

The boreal ecosystem stretches across the Northern Hemisphere’s circumpolar countries. It covers approximately 10% of the Earth’s land surface, ranking second in terms of total plant mass to the tropical forest belt. Because it contains approximately 40% of terrestrial carbon, it plays an extremely important role in the global carbon budget. It is therefore important to identify anything that perturbs this ecosystem, for example, fire. However, global terrestrial carbon cycle models generally do not take into account loss of carbon through disturbance. Further, disturbance has a strong influence on succession in the boreal ecosystem through the effect on opportunities for change from boreal biotopes to others typical of biomes to the south and north. Since climate change may alter the frequency and size of disturbance events, it is vital to monitor changes in their spatial and temporal occurrence if we are to predict the impacts of global environmental change. Yet, currently there is no comprehensive database of disturbance across the entire boreal ecosystem and, where efforts to collate information have been attempted, they have usually been either spatially restricted or a snapshot in time. Air photo interpretation and visual annotation of base maps from light aircraft are the primary methods for fire mapping. This is an extremely labour intensive method of fire scar mapping which requires considerable financial investment. For the boreal forest, the large area of individual burns makes coarse resolution remote sensing an attractive alternative although it is limited to the last 25 years.

This paper extends the work of Eastwood at al. (1998) on fire scar detection through the comparison of new indices, thresholding and segmentation of VEGETATION data over the BOREAS experimental region. Imagery on a monthly time-step was acquired through the 1998 active fire season (May-September). Segmentation was performed on a total of 24 VEGETATION images to assess sensitivity of the approach to segmentation criteria and the variability of fire scar detection as a function of image geometry and atmospheric state. The results were compared against hot spot observations recorded for the time period in the FIRE-M3 detection system () (Li et al. 1997). Older fire scars were identified with reference to the Canadian Forestry Service GIS database of large fires covering the period 1980-92 as used by Eastwood et al. (1998). The results are in accordance with the observations by Eastwood at al. (1998) that the middle infrared waveband provides better spectral differentiation of fire scars than methods based on NDVI.

Monitoring Boreal Forest Resources in Northern Europe from the VEGETATION instrument

Bernard Pinty(1),

Jean-Luc Widlowski(1), Nadine Gobron(1), Michel M. Verstraete(1), Ola Engelsen(2), Harald Johnsen(2) and Yves Govaerts(3)

(1) Space Applications Institute

EC Joint Research Centre

I--21020 Ispra (VA), Italy

(2) Norut Information Technology

N-9005 Tromso, Norway

(3) EUMETSAT, Am Kavalleriesand 31

D-64295 Darmstadt, Germany

Boreal forests constitute a major renewable economic resource of north European countries. These ecosystems are threatened by the combined effects of increasing demand for timber and wood products, and by likely changes of climate resulting from the well known increase in greenhouse gases in the atmosphere. The VEGETATION sensor offers new opportunities to monitor the status and evolution of these resources, thanks to its four spectral bands. The synergy between these data and the existence of advanced models and techniques of data analysis has permitted the full exploitation of data generated by this instrument, despite the difficulties arising from low solar zenith angles and a variable atmospheric aerosol load. This project provides a map showing the likelihood of the presence of dense boreal forest in the Barents region. This information should be useful for the pre-operational management of northern Europe's boreal forest resources. The approach developed here will also be applicable to the operational management of other ecosystems.

SPATEM: The analysis of annual sequences of VEGETATION data at the landscape scale.

Agustin Lobo and Nicolau Pineda

Instituto de Ciencias de la Tierra "Jaume Almera" (CSIC)

Lluis Solé Sabarís s/n, 08028 Barcelona, Spain

alobo@ija.csic.es

The instrument VEGETATION acquires images with global and daily coverage of the Earth surface at 1 km2 resolution from the same SPOT platform that hosts HRVIR, the sensor acquiring high spatial resolution images of selected areas. VEGETATION and HRVIR imagery have similar spectral characteristics, are acquired with similar angles and can be geo-corrected to the same projections. In SPATEM we have approached the integration of multi-temporal VEGETATION images into products derived from the high spatial resolution images in a forested Mediterranean landscape. Our approach has included both methodological aspects as well as the implications of such integration for the applications of Earth Observation.

We used the high spatial resolution imagery and a method based on image segmentation and discriminant analysis to produce a detailed land-cover map, emphasizing on forest types, and we calculated the cover fractions of the classes within each VEGETATION pixel. After a general analysis of the geometric and optical quality of the multi-temporal sequence of VEGETATION images and an optimization of the annual sequence of vegetation indices, we modeled the annual cycle of the vegetation indices of each class. We also performed an inverse analysis, in which we estimated the cover fractions from the multi-temporal sequences of the VEGETATION pixels and the models, using an Spectral Mixture Analysis. Results were poor at the pixel level, probably as a result of our models not being at extremes in the feature space, but the average estimate for the entire area of study was very accurate. We discuss some possible improvements to increase accuracy at the 1 km2 pixel level.

We also discuss the characteristics of the different annual cycles considering both the ecological attributes of the classes and their projections on a reduced plane of phenologic variability. In order to obtain a more comprehensive view of different annual cycles, we also enlarged our area to cover 17 digital vegetation maps at 1:50 000 scale within the area covered by our VEGETATION imagery (0 to 5 E, 40 to 45 N). In this case, we selected those 1 km2 pixels that were at least 90% contained within one single type of vegetation according to the maps. We found that VEGETATION data can define distinct annual cycles of vegetation indices for very detailed vegetation types. This ability will have important consequences, not only for improving global-scale land-cover mapping, but also because a better understanding of the annual cycles will let us formulate and test improved models of vegetation dynamics, beyond the elementary vegetation units that are currently used.

VEGETATION/SPOT4 applications for macro-regional landscape mapping

(Land Cover/ Land Use mapping and monitoring of Russia, PI - E.V. Milanova)

Lioubimtseva E. (1),

E.V Milanova (2), P.Tcherkashin(2), and V.N.Solntsev (2)

(1) Dept. of Geography and Land Studies, Central Washington University

(2) Dept. of Physical Geography and Geoecology, Moscow Lomonosov State University

The final (post-launch) phase of this study has been focused on analysis of the potential of VEGETATION/SPOT 4 for medium scale land-cover mapping and, more specifically for mapping and landscape mosaics of different heterogeneity and graininess. The area of study comprises various combinations of coniferous and mixed forests with forest-steppe, pastures and agricultural lands on the Russian plain. During the preparatory phase data from AVHRR/ NOAA, RESURS01-3/MSU-SK, and COSMOS/MK-4 were processed to explore the requirements of landscape ecological mapping in terms of spectral channels and ground resolution of the satellite instrument.

Evaluation of 16 most common types of landscape mosaics identified within the study area demonstrated that despite its relatively coarse ground resolution, VEGETATION provides an excellent tool not only for land-use/ cover mapping but also, and especially, for landscape ecological applications. Available spectral windows allow particularly fine discrimination between patches within complex heterogeneous landscapes (i.e. forest-steppe mosaics with agricultural lands) as well as discrimination between different types of forest vegetation. The data also represent considerable interest for monitoring riparian ecosystems and water content analyses in vegetation cover.

Fire patches in natural vegetation in southern Africa

Swinnen E. *, Verwimp R. **, Gulinck H. ***

* KULeuven / VITO (since 15/9/1999)

** Ground for GIS, KULeuven

*** KULeuven

The objective of this study is to explore the feasibility of SPOT-VEGETATION for burnt surface mapping in savanna ecosystems. The study site is located in the Chobe area, northern Botswana. It is mainly covered by a mosaic of open and dense woody savanna and woodland. A time series with an interval of 2 weeks of VEGETATION images over one dry season (1998) is used for the signature analysis and the change monitoring. End member analysis is applied on a single date image with aid of a high resolution SPOT-XS image of even date and classified by means of field survey recordings. The spectral behaviour of burnt and unburnt savanna is examined in the RED, NIR and SWIR bands on a single image and for a time series. The reflectance of a burnt surface is lower than for savanna: a decrease in reflectance of 25-30% in NIR and 15-20% in SWIR was observed immediately after burning. The signal variability is largest in SWIR compared to NIR and RED. An increase of variability is observed in all examined bands immediately after a fire occurred. NIR has the most discriminative power. The change detection analysis focussed on mapping fire scars (change) and savanna (no change). Two change algorithms are employed to extract the change information: standardised differencing and 2-dimensional principal component analysis (PCA) (time1 vs time2). A statistical threshold is applied to classify the output of these algorithms to a burnt/non-burnt information layer. Accuracy is assessed using the field work recordings. Both change algorithms are successfully applied. All kappa coefficients are higher than 80%. No significant difference can be proved between standardised differencing and PCA. Although the slower decrease of the SWIR band after the burn-event the change monitoring process performed no significant better results for NIR. For areal extent measurement, two sub-pixel classifiers are examined, linear spectral unmixing (LSU) and artificial neural networks (ANN). The former is based on typical class signatures, but omitting the spectral heterogeneity within one class for a given band, whereas the latter takes this heterogeneity into account by training the network with examples. Different band combinations are explored. Results are evaluated by comparing the areal estimates of the classes, the rank correlation coefficient with the reference high resolution classification and the distributions of the errors. Both methods give satisfactory results, although ANN performs slightly better. Band combinations including NIR always yield better results. When SWIR is added, the results are less accurate for LSU, because of the large variability of the reflectance of both classes in this spectrum. The best combination is RED-NIR for both techniques. Rank correlations obtained range between .85 and .90 for LSU and exceed .90 for ANN. Largest estimation errors are logically found for the mixed pixels. The result of ANN contains more small errors than LSU (resp. 75% and 35% of the pixels with an error ................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download