Updated: 11th Feb. 2009
![[The University of Leicester]](siteimages/uni100.gif)
Mark Doherty - Keynote
Head of Earth Observation, Exploitation and Services Division
European Space Agency
Title: ESA earth observation programmes and climate change
When: Tuesday, 8/Sep/2009: 1:15pm - 2:15pm
Abstract: Mark will present an overview of the ESA Climate Change Initative for which ESA will be issuing ITTs in November 2009. Much of the presentation will cover issues related to 'Essential Climate Variables' and the ESA contribution to mapping and monitoring these variables.
Chris Justice - Keynote
University of Maryland, NASA MODIS science team and GOFC-GOLD member
Title: Global fire: a remote sensing research and applications perspective
When: Tuesday, 8/Sep/2009: 2:15pm - 3:15pm
Abstract: The view from space provided by global earth-observation satellites show the Earth as a Fire Planet. Fire has helped shape the current distribution of global vegetation and in many ecosystems continues as a periodic disturbance. Arguably fire was even more prevalent in former times, before the introduction of modern methods of fire management and suppression. However, catastrophic wildfires particularly at the urban-wildland interface, remind us of the complexity and importance of managing fire. Fires have a local impact on nutrient cycling and water and air quality at regional scales. As an important land-surface process, fires present some interesting research challenges for Earth System Science. For example, trace gas emissions from fires contribute to global warming, fires alter the land-surface, removing vegetation and altering the surface albedo and there are important interactions between smoke and cloud formation. There has been a long relationship between human land-use and fire. Fire is still used widely as an inexpensive land management tool in many parts of the World, for example to clear the land, maintain pastures or remove crop residue. In several areas traditional fire management practices are changing with population growth and dynamics for, example as land is abandoned or new areas are developed. Fire regimes are also changing with the increasing frequency of extreme weather events and a warming climate. Understanding the underlying processes fire weather, fuel, ignition and the human use of fire, will be important for developing fire management procedures in the context of reducing global greenhouse gas emissions or developing realistic projections of how fire regimes may change in the future and how the negative aspects of such changes in fire regimes might be mitigated. Various satellite systems with different sensing capabilities are being used to research different aspects of fire. As a result of remote sensing research, the processing and analysis procedures have been sufficiently well-developed, that satellite fire data are now being used in practical resource management applications. With current coarse resolution satellite observations we are now able to detect, characterize and monitor fires globally, over several years. The NASA MODIS and Landsat instruments provide important tools for developing such a long-term record. The continuity of data from such systems will be important, if we are to establish a long-term data record with which to quantify how fire regimes are changing and to continue to provide data to the operational fire monitoring systems. The satellite fire research and user communities through the international Global Observations of Forest Cover and Landcover Dynamics (GOFC-GOLD) program are working to improve coordination and cooperation on fire observations, encourage the development of new sensing capabilities, improve data coverage and availability and ensure the continuity of long-term satellite fire observations.
Dirk Rieke-Zapp - Invited Speaker
Institute of Geological Sciences, University of Bern
Title: The application of photogrammetry for process quantification in geomorphology – examples of modern applications
When: Thursday, 10/Sep/2009: 8:45am - 10:15am
Abstract: Photogrammetry is a well established tool in the earth sciences for retrieval of qualitative as well as quantitative information from imagery. In the past, analysis was predominantly performed on publicly available or archival imagery taken by trained photogrammetrists. In this case mission planning, image acquisition, camera calibration, and image orientation were often readily available to earth scientists at the start of their analyses. The photogrammetric products that are of interest to earth scientists are typically digital elevation models (DEMs) or orthophotos that are processed for analysis. Monitoring of changes in the landscape, or tracking the course of a laboratory experiment requires data for different time steps for quantification of processes. Although earth scientists apply well-established photogrammetric techniques for data acquisition, workflow and hardware setup can pose challenges to the users with little experience. Expectations regarding accuracy in object space are often close to the theoretical limit of a given setup and are difficult to meet under field conditions or during short breaks of experiments for data capture. Therefore, the workflow for data acquisition must be streamlined and the hardware optimized to yield optimum accuracy under these conditions. In this paper we will present some geomorphologic experiments where digital photogrammetry was employed for data capture. Special attention was given to uncommon or novel solutions of camera setup and hardware easing data capture or image processing for analysis in the earth sciences
Wolfgang Wagner - Invited Speaker
Institute of Photogrammetry and Remote Sensing
Vienna University of Technology
Title: Terrain characterization and vegetation structural analysis with full-waveform airborne laser scanners
When: Tuesday, 08/Sep/2009: 4:15pm - 5:45pm
Abstract: Airborne laser scanning (ALS) is an active remote sensing technique that uses lasers to transmit nanosecond-long laser pulses with a high pulse repetition frequency. Over vegetated areas the laser pulses may be reflected by the leaves and branches of the vegetation and the underlying terrain surface. Given laser footprint sizes in the order to 0.2-1.0 m the laser pulses may be reflected by several discrete objects, potentially yielding relatively complex backscatter signals referred to as "waveforms". By registering the full-waveform during the flight and by decomposing it into a series of echoes in a post-processing step, information about the geometric location and radiometric (scattering) properties of each individual scatterer can be obtained. Even though research on small-footprint waveform data can still be considered to be only in its beginning, a number of benefits start to emerge. From a theoretical point of view, the value of full-waveform ALS sensors lies in the fact that the measurement process is depicted in its entire complexity. Thus it is possible to physically model the measurements process. From a practical point of view, the combination of geometric and radiometric information offers new means for interpreting the derived 3D point cloud and for deriving vegetation and terrain information. In this contribution we discuss methods for processing full-waveform ALS data for improving the characterization of the terrain and vegetation structure. The results show that full-waveform information is particularly useful for characterizing low vegetation cover, terrain roughness, and different forest classes. Examples are drawn from study areas in eastern Austria.
Tristan Quaife - Invited Speaker
Department of Geography
UCL
Title: Equivalence of data assimilation techniques in land remote sensing problems.
When: Wednesday, 09/Sep/2009: 4:00pm - 5:30pm.
Abstract: Data assimilation (DA) refers to a set of mathematical techniques that seek to combine models and data in a statistically optimal fashion. Such methods are receiving growing interest within the land remote sensing community as a means integrating remote sensing observations with land surface process models. Partly this level of interest is due to the success of DA in the atmospheric modelling community where satellite data is assimilated into atmospheric transport models on an operational basis. The functioning of a DA system is dependent on accurate characterisation of error statistics for the observations and the model that the data is being assimilated into. In the case of the model this is often a non-trivial problem and it is common to resort to the use of a diagonal covariance matrix populated by crude estimates of the uncertainties in various model components; correlations between the components are generally ignored. Furthermore, traditional DA systems implicitly assume that the underlying model can be described by a first order Markov process, which is appropriate for atmospheric problems but not so for the land surface where temporal correlations will be strong. With reference to a newly developed technique for retrieving surface BRDF parameters, based on the original MODIS BRDF/albedo algorithm, this paper explores various DA techniques and demonstrates the equivalences between them with reference to Bayes theorem and constrained linear inversion. The model explored is a constraint of smoothness on parameter retrieval. This is particularly appropriate to the land surface problem where observations on one day should not depart greatly from those on the previous day. Results are provided in the form of temporal profiles of BRDF model parameters derived from MODIS data.