Soutenance mémoire
In this introductory chapter the objectives of the thesis are put in perspective. The need and current initiatives for global monitoring of vegetation are first discussed. The significance and challenges associated with the estimation of LAI, specifically from remote sensing observations for global monitoring are briefly dealt. Methods for the estimation of LAI as well as temporal filtering methods for smoothing and gap filling satellite estimates are introduced. Finally, the specific objectives of the thesis are outlined along with the structure of the document.
Global monitoring of surface vegetation
Global monitoring of vegetation variables is of importance for the description and understanding earth dynamic processes, particularly in the context of the on-going global climate change to design policies for ensuring food security, conserving biodiversity and mitigate climate change effects through efficient management of resources. Remote Sensing observations from satellites, particularly from medium resolution satellite sensors, provide global observations over long-time periods. These observations would provide the policy makers of the different countries with the information that would help them to make their decisions. Consequently, vegetation monitoring at global and regional scales plays an important role in the economic development and well-being of the nations. Many initiatives have been taken over the years to enable continuous and consistent observations to the users. Some of the notable initiatives are:
1. Global Climate Observing System, GCOS (GCOS, 2012), established in 1992, is a joint initiative by the World Meteorological Organization (WMO), the Intergovernmental Oceanographic Commission (IOC) of the United Nations Educational Scientific and Cultural Organization (UNESCO), the United Nations Environment Program (UNEP) and the International Council for Science (ISCU).GCOS aims to provide comprehensive information pertaining to earth climate processes for long-time periods. Information on vegetation and other terrestrial process are provided by the Global Terrestrial Observing System (GTOS, 2007).
2. Group on Earth observations, GEO (GEO, 2012) is a voluntary partnership of governments and international organizations to provide a global framework for projects and efforts aimed at providing Earth observations. GEO is coordinating the international effort to build Global Earth Observation System of Systems (GEOSS), a global public infrastructure for Earth observations consisting of flexible and distributed network of content providers. This framework was established in 2005, in response to the calls for sustainable development of G8 countries. As of January 2008, 72 Governments and the European Commission have been participating in this initiative. In addition, 52 intergovernmental, international and regional organizations with a mandate in Earth observation have also been recognized as Participating Organizations. GEO also coordinates the “GEO-GLAM” initiative, for agricultural monitoring at the continental and global scales.
3. Global Monitoring for Environment and Security, GMES (GMES, 2012) is a joint initiative by the European Commission and European Space Agency with an attempt at creating an autonomous and operational Earth observation capacity from multiple-sources. The main users of the GMES initiative are the policymakers. The services of GMES can broadly be classified into: land, marine and atmospheric services, emergency and security services, and climate change services. The GMES uses its satellites and non-space assets to provide services to its users. Geoland 2, the context of project comes under the land, marine and atmospheric services of the GMES.
GEOLAND2(Geoland2, 2011a), a GMES services framework for land monitoring, aims at building user-driven operational land monitoring services. It derives from earlier projects of the European Commission and European Space Agency for the development of time- and cost-effective geo-information. It realizes its objectives through two of its layers, namely, the core mapping services and core information services (Geoland2, 2011b).
• The Core Mapping Services produce the basic earth observations on land-cover, land-use, the annual and seasonal vegetation cycles, radiation budget and water cycles from satellite observations.
• The Core Information Services are a set of thematic elements built upon the Core Mapping Services to generate products pertaining to European Environmental Policies and International treaties. Core Information Services also compares and evaluates the products of Core Mapping Services with products from other common approaches from an operational perspective. The Core Information Services include spatial planning, Agri-Environment Monitoring, Water Monitoring, Forest Monitoring, Land Carbon assessment, Natural Resource Monitoring in Africa and Global Crop Monitoring.
Leaf Area Index in Ecological perspective
Leaf Area Index (LAI), is a biophysical variable accessible from remote sensing and identified to be essential for monitoring the earth climate processes (GCOS, 2010). LAI is defined as half the area of the green elements per unit horizontal area (Chen and Black, 1992). It provides an estimate of the mass-energy interface between the surface vegetation and the atmosphere. Thus, it is of interest to many studies including climate change, global carbon fluxes, land cover and crop monitoring.
LAI is one of the primary canopy variables that control the photosynthetic activity in the canopy. LAI along with canopy clumping index was found to be an important parameter for modeling plant growth and carbon cycle (Chen et al., 2003). LAI directly influences the photosynthetic capacity of the canopies and consequently, is a primary biophysical control on the net ecosystem productivity (Barr et al., 2004). LAI controls fAPAR, i.e. the fraction of photosynthetically active radiation (400-700 nm) absorbed by the canopy. The big-leaf model used for the canopy photosynthesis estimation requires empirical curvature factors in response to irradiance. This curvature factor, for any given distribution of leaf nitrogen content in the canopy, is dependent on LAI and leaf nitrogen content (De Pury and Farquhar, 1997). This observation led (De Pury and Farquhar, 1997) to treat the sunlit and shaded leaves separately to obtain improved estimates of photosynthesis as compared to using fAPAR. Further, from studies on Maize canopies the LAI is found to be predominant factor for describing the canopy processes where water stress varies gradually over long periods (Steduto and Hsiao, 1998).
Estimation of LAI
LAI can be estimated from ground measurements based on direct or indirect methods (Weiss et al., 2004;Jonckheere et al., 2004). Direct methods like area harvesting for destructive sampling, using allometric relations and litter traps are measurement intensive. Indirect methods like LICOR-2000, TRAC or hemispherical photography (‘CanEye’) use the penetration of light through the canopies to estimate the LAI (Lang and McMurtrie, 1992). Although much efficient than destructive methods, they still need intensive ground sampling in case of coverage of large domains. Therefore, groundbased methods cannot be used for estimations at global scale or for long periods due to economic and labor constraints.
Remote sensing observations from satellite sensors provide an opportunity to estimate LAI on a global scale for long periods at high revisit frequency. Medium resolution satellite sensors like MODIS (Knyazikhin et al., 1999), AVHRR (Baret et al., 2011a) and VEGETATION (Baret et al., 2007;Weiss et al., 2007a) provide long term estimation of LAI. Canopy reflectance at various wavelengths observed by the sensors is then used to retrieve LAI. Canopy reflectance results from the radiative transfer processes within the canopy and includes the contribution from the soil background. While forward modeling establishes a causal relation between the biophysical variable of interest, LAI, with the observed reflectance measurements the retrieval is an inversion problem, which is usually ill-posed (Combal et al., 2003), i.e. several possible solutions can correspond to a given set of measurements.
|
Table des matières
Acknowledgements
Abstract
1 Introduction
1.1 Global monitoring of surface vegetation
1.1.1 Leaf Area Index in Ecological perspective
1.2 Estimation of LAI
1.2.1 The biophysical variable driven methods
1.2.2 The radiometric data driven methods
1.3 Time Series Estimates of LAI
1.4 Objectives and structure of the Thesis
2 Inversion of Spectral Reflectance for the estimation of biophysical variables On the Effective Nature of LAI as Observed From Remote Sensing Observations
3 Gap-filling and smoothing of MODIS LAI time Series data
A comparison of methods for smoothing and gap filling time series of remote
sensing observations: application to MODIS LAI products
4 Near-Real Time Estimation of LAI from daily AVHRR LAI estimates
Near real time estimates of Leaf Area Index from AVHRR time series data
5 Conclusions and Possible Future Work
5.1.1 Effective Nature of LAI Estimations from Remote Sensing observations
5.1.2 Improving the quality of MODIS time series estimates
5.1.3 Near Real-Time Estimation of LAI from AVHRR daily estimates of LAI
References
Télécharger le rapport complet
