• Aucun résultat trouvé

Understanding changes in the landscape based on a Landsat remote sensing analysis in the Karbi Anglong hills, Assam, India

N/A
N/A
Protected

Academic year: 2021

Partager "Understanding changes in the landscape based on a Landsat remote sensing analysis in the Karbi Anglong hills, Assam, India"

Copied!
2
0
0

Texte intégral

(1)

CONGRESS OFFICE

23/25, rue Notre-Dame des Victoires 75002 Paris | France www.hopscotchcongres.com

PROGRAM

&

ABSTRACTS

(2)

Le Corum, Montpellier, 19-23 June 2016

324

|

53

rd

ATBC 2016

A

BSTRACTS /

ORAL PRESENTATIONS

O58-05 – S58 Monitoring and mapping tropical biodiversity and ecosystem services with remote sensing

Thursday 23 June / 10:30-12:00 – Sully1

Analysis of the potential of 10m resolution optical imagery for the monitoring of logging impacts

on the forest cover in the Republic of Congo

VERHEGGHEN ASTRID, EVA HUGH, ACHARD FRÉDÉRIC

European Commission, Joint Research Centre, Institute for Environment and Sustainability, 21027, Ispra, Italy

Background: Large areas of the African moist forests are being logged and little is known about the impacts on the forest and the related carbon emissions (Gourlet-Fleury et al. 2013). While logging roads can be persistent over time and captured in remote sensing derived products such as the Global Forest Cover Change (Hansen et al. 2013), the dynamic of associated openings in the canopy cover caused by skid paths, log decks and tree removal are not well documented due to temporal and spatial limitations of satellite images. The potential of the recently launched Sentinel-2a satellite to overcome those limitations and monitor small scale disturbances due to selective logging is investigated in this study.

Method: A unique time series of 10 m resolution satellite image is assembled from images acquired during the SPOT5-Take5 experiment complemented by the first Sentinel-2a acquisitions to monitor a certified selective logging concession in the Northern part of the Republic of Congo during the year 2015. Vegetation indices and a spectral un-mixing model are applied to the images in order to highlight gaps in the canopy. The satellite dataset is used to assess both the location and timing of logging activities by mapping the extent of canopy cover changes. The vegetation regeneration dynamics after logging events are assessed thanks to the high temporal frequency of image acquisitions.

Result and Discussion: Such imagery time series is suitable to monitor the temporal dynamics of logging roads and small disturbances in the forest cover related to wood extraction. We demonstrate that small openings in the forest canopy disappear in just a few months, underlining the requirement for high image acquisition frequency as provided by the Sentinel-2 mission. The resulting map of selective logged areas will be combined with field data to assess the area impacted on the ground. We will also investigate the possibility of linking such maps to measurements of wood extraction volume in the view of estimating changes in the Carbon stocks of logged forests over extended areas.

Additionally, the networks of logging roads will be analysed using a morphological spatial pattern software to compare the intensity of logging activities between this concession and a non-certified logging concession.

O58-06 – S58 Monitoring and mapping tropical biodiversity and ecosystem services with remote sensing

Thursday 23 June / 10:30-12:00 – Sully1

Bioclimatic envelope models predict a decrease in tropical forest carbon stocks with climate

change in Madagascar.

GHISLAIN VIEILLEDENT

1

, OLIVER GARDI

2

, CLOVIS GRINAND

3

, CHRISTIAN BURREN

4

,

MAMITIANA ANDRIAMANJATO

5

, CHRISTIAN CAMARA

6

, CHARLIE J. GARDNER

7

, LEAH GLASS

8

,

ANDRIAMBOLANTSOA RASOLOHERY

9

, HARIFIDY RAKOTO RATSIMBA

10

, VALÉRY GOND

11

,

JEAN-ROGER RAKOTOARIJAONA

12

1JRC, Forest Resources and Climate, 21027, Ispra, Italy

2Bern University of Applied Sciences, HAFL, 3052, Zollikofen, Switzerland 3ETC Terra, ETC Lab, 5020, Paris, France

4WCS, Madagascar programme, 101, Antananarivo, Madagascar

5Ministère de l’Environnement et des Forêts, Direction Générale des Forêts, 101, Antananarivo, Madagascar 6Missouri Botanical Garden, Madagascar programme, 101, Antananarivo, Madagascar

7WWF, Madagascar and Western Indian Ocean Programme Office, 101, Antananarivo, Madagascar 8Blue Ventures, Blue Forests program, 101, Antananarivo, Madagascar

9Conservation International, Africa and Madagascar Field Division, 101, Antananarivo, Madagascar 10Université d’Antananarivo, Département des Eaux et Forêts, 101, Antananarivo, Madagascar 11Cirad, UPR BSEF, 101, Montpellier, France

12ONE, Direction des Informations Environnementales, 101, Antananarivo, Madagascar

Recent studies have underlined the importance of climatic variables in determining tree height and biomass in tropical forests. Nonetheless, the effects of climate on tropical forest carbon stocks remain uncertain. In particular, the application of process-based dynamic global vegetation models have led to contrasting conclusions regarding the potential impact of climate change on tropical forest carbon storage.

Using a correlative approach based on a bioclimatic envelope model and data from 1771 forest plots inventoried during the period 1996-2013 in Madagascar over a large climatic gradient, we show that temperature seasonality, annual precipitation and mean annual temperature are key variables in determining forest above-ground carbon density.

Taking into account the explicative climate variables, we obtained an accurate (R2 = 70% and RMSE = 40 Mg.ha-1) forest carbon map for Madagascar at 250 m resolution for the year 2010. This national map was more accurate than previously published global carbon maps (R 2 < 26% and RMSE > 63 Mg.ha-1).

Combining our model with the climatic projections for Madagascar from 7 IPCC CMIP5 global climate models following the RCP 8.5, we forecast an average forest carbon stock loss of 17% (range: 7-24%) by the year 2080. For comparison, a spatially homogeneous deforestation of 0.5% per year on the same period would lead to a loss of 30% of the forest carbon stock.

Our study shows that climate change is likely to induce a decrease of tropical forest carbon stocks. This loss could be due to a decrease in the average tree size and to shifts in tree species distribution, with the selection of small-statured species. In Madagascar, climate-induced carbon emissions might be, at least, of the same order of magnitude as emissions associated to anthropogenic deforestation.

O58-03 – S58 Monitoring and mapping tropical biodiversity and ecosystem services with remote sensing

Thursday 23 June / 10:30-12:00 – Sully1

Characterizing texture- structure relationship in the tropical forests of Western Ghats of India

using high resolution Cartosat Imagery

SOURABH PARGAL

1

, PIERRE COUTERON

1

, RAPHAEL PELISSIER

1

, RAKESH FARARODA

2

,

GOPALAKRISHNAN RAJASHEKHAR

2

, CHANDRA SHEKHAR JHA

2

1Institut de Recherche pour le Développement (IRD), UMR AMAP (botAnique et Modélisation de l’Architecture des Plantes et des

vegetations), 34398, Montpellier, France

2NRSC (National Remote Sensing Center), Forest and Ecology Division, 500625, Hyderabad, India

Context: Regional assessment of forest aboveground biomass (AGB) in tropical forests is spoilt by uncertainty. There is thus a major challenge in providing accurate carbon stock assessments especially in the context of REDD+ program. One of the major reasons for uncertainty in AGB estimates is the saturation of remote sensing signals at relatively low AGB values, for both optical reflectance and radar backscattering signatures. Canopy texture analysis of very high resolution (VHR) imagery have provided encouraging results for retrieving forest stand structure parameters for different tropical forests in Central Africa, French Guiana and India, with AGB values going above 650 Mg/ha. The method, Fourier Transform Texture Ordination (FOTO), performs well in characterizing texture-structure relationships and quantifying AGB variations, when tested for areas with a given forest type. However the robustness of such approaches across different forest types characterized by different forest structure and allometries, is more demanding and yet under investigation.

Method: In this study we focused on a gradient of tropical forest types, in the Western Ghats of India, ranging from dry deciduous to wet evergreen forests, in the Yellapur division, Uttara Kannada, Karnataka, India. Canopy texture analysis was done using the FOTO method in order to characterize the texture gradient present in the study area using Cartosat-1a imagery, a 2.5 m spatial resolution panchromatic sensor (500 to 850 nm) launched by Indian Space Research Organization (IRSO). We established 14 1-ha forest plots covering the whole gradient of canopy texture and forest types encountered in the study area and forest structural data was obtained. These 14 plots were used to calibrate texture - structure model for the forest types present in the area.

Result: Correlation between observed and predicted AGB was found to be good (R2 = 0.83) for nine plots lying in wet evergreen and moist deciduous zone which both displayed closed canopy, however the correlation dropped when four plots lying in fairly open canopy in dry deciduous or dry to moist transition zone were added to the analysis.

Conclusion: The FOTO method performed well in characterizing texture-structure relationships even with strong gradients present in the study area. However further investigations would be needed to test a larger gradient in forest types, such as including fairly open canopy forests in the dry deciduous region.

O58-04 – S58 Monitoring and mapping tropical biodiversity and ecosystem services with remote sensing

Thursday 23 June / 10:30-12:00 – Sully1

Understanding changes in the landscape based on a Landsat remote sensing analysis in the

Karbi Anglong hills, Assam, India

TOBIAS SCHMID

1

, SWEN BOS

1

, JOHAN OSZWALD

2

, VALERY GOND

3

, CLAUDE GARICA

3

1ETH Zürich, Environmental Sciences, 8092, Zürich, Switzerland 2Université de Rennes 2, COSTEL, 35043, Rennes, France 3CIRAD, ES, 34398, Montpellier, France

The landscape of the Karbi Anlong hills (State of Assam, India), south of the Kaziranga National Park, is shaped by small-scale farmers of the Karbi tribe. They traditionally practice jhum cultivation of upland rice and have started to cultivate cash crops such as bamboo, tea and rubber to improve their livelihoods. The forests of the Karbi Anglong hills also provide a crucial habitat for many flagship species, such as the Rhinoceros (Rhinoceros unicornis) and Tiger (Panthera tigris) during the monsoon months, while the Brahmaputra river floods the plains of Kaziranga.

Analyzing the historical changes in the landscape is a necessary first step to understand the forces driving land use and land cover change in the Karbi Anglong ecosystem. This information can then be used to identify practices, understand drivers and then design management interventions and policies, as part of an integrated landscape approach.

The forests of Karbi Anglong were analyzed through a GIS analysis of Landsat images from 1988 to 2016. Prior to classifying the forests, a succession and landscape dynamic model of this region was designed. Then a supervised classification was conducted throughout the northern Karbi Anglong hills to gain a full understanding of the forest structure and composition.

Human influence has shaped the landscape of the northern Karbi Anglong hills, whereas of lately an extensification was observed, which qualitative interviews attribute to political instability. This observation could be proven by the analysis of the Landsat (5&8) imagery as large proportions of young succession and immature forests were found throughout the study area. In addition fewer young yum fields were found within the study area.

Within this study we were able to establish a first result on landscape change that will serve as the foundation for future work in developing a landscape approach in the Karbi Anglong hills.

Références

Documents relatifs

The influence of spatial resolution on the sensitivity of r-spectra to capture canopy grain of different forest types was highlighted using 1-m panchromatic and 4-m near infrared

1) Building the search space: In order to implement these 2 algorithms, a representation of the search space under the form of a meta-graph has been built. Each node of this new

The method is tested with multi-spectral IKONOS images acquired on areas containing six major communities (oak, pine, fir, primary and secondary high tropical forests, and

™ Results : The 500-m landscape closure index was positively correlated with higher serological incidence (p &lt; 0.05).. Contact: *soti@teledetection.fr *vs.nev@ntropic.fr ) ) ) ) )

Then, landscape variables were defined for each pond in the area based on bibliographic knowledge of the vector ecology: a landscape closure index calculated within a 100, 500 and

In this paper we designed, developed and validated a novel framework which address concurrently the registra- tion and change detection tasks in very high resolution mul-

Identifying degraded forests in an Amazonian landscape

a  Researcher, CIRAD, UMR TETIS F‐34398, Montpellier, France,  b  PhD Student, presently at UMR Espace‐Dev, Université de la Réunion,