Click here to print this project Click here to print this project|
Status: Confirmed |
Open to sharing: Yes |
Confidential: No |
|
Transnational Access: Yes |
Open to training: No |
Grounded / Maintenance: No |
Aircraft name: CASA 212 RS - INTA
Airport: For this study we selected three cities along an urban density gradient in Flanders, Belgium: (i) the Brussels Capital Region (population > 1 million), (ii) Ghent (population > 250.000) and (iii) Leuven (population < 100.000). Sampling along a density gradient for three different study sites will allow us to evaluate the robustness of our remote sensing derived land cover products with respect to urban density and complexity. The Brussels study area has been the object of study of extensive hydro-ecological and remote sensing research during the past years and is currently the key study area within the UrbanEARS project. Consequently a lot of field measurements (and experience) have been carried out on that study area which may be useful for calibration and validation purposes within this project. Brussels, as well as the other two selected cities are also the host cities of the three universities involved in this proposal. This will not only facilitate the practical organization and intensity of field campaigns, but will further allow us to broaden the exploitation of the requested data for both research and educational purposes. For students it is not only more appealing to work on data of an area they properly know, the requested AHS data will also allow students to easily verify in the field there results obtained during practical sessions of remote sensing master courses (see also section "Other useful comments"). In addition, master and bachelor thesis projects can more easily be facilitated and supported this way.
Project theme: TA-005 Airborne imaging for environmental science applications
Science context: Although remote sensing is an interesting source for characterising the physical properties of the urban environment, in terms of hydrological or urban heat flux modelling the level of differentiation in functional land cover properties delivered by conventional remote sensing approaches [1] is rather coarse and does not allow a clear parameterisation of pixel components in terms of their biophysical behaviour (e.g. level of permeability, thermal properties). The data requested here will be used to explore the potential of the combined use of hyperspectral, thermal and LiDAR sensing technologies for detailed characterization of urban green and the built environment. We will specifically focus on evaluating the added value of data fusion approaches in solving pending issues such as: *the limitations of current approaches to address problems related to spatial ambiguity and similarity between materials and disturbing effects of illumination and viewing geometry [2]-[6] *the optimisation of spectral unmixing techniques for detailed biophysical characterisation of urban areas, and the transferability of these approaches to coarser resolutions. This is essential, particularly in the context of new hyperspectral missions like EnMAP [7] and HyspIRI [8], which will open up opportunities for operational use of hyperspectral remote sensing in urban ecosystem analysis at the metropolitan scale. The work to be carried out will be closely linked to the ongoing BELSPO project UrbanEARS (Urban Ecosystem Analysis supported by Remote Sensing) which will enable us to directly use the outcome of the experiments planned as an input for urban heat island modelling and modelling of urban hydrology.
Measurements to be made by aircraft: This experiment aims at exploring the potential of the combined use of hyperspectral visible, near-, shortwave- and thermal infrared spectral information, obtained from the AHS sensor, in combination with structural information derived by LiDAR (obtained thought the Flanders Geographical Information Agency), for detailed characterisation of morphological and (bio)physical properties of the urban environment, relevant for heat and water regulation related research. We strive to provide universal insights and develop generic methodologies. Therefore data are requested for three urban areas along an urban density gradient. In order to allow for a comprehensive modelling of relevant spatial extent, an emphasis in the methodological framework will be put on upscaling, including analysis of the potential of data from future satellite programs such as EnMAP, HyspIRI and Sentinel2 for urban environmental modelling. The latter will be simulated from the airborne AHS data. More specifically the research objectives can be detailed as follows. Through data fusion approaches we will: - Explore the potential of innovative spectral unmixing approaches for improved land cover parameterisation of urban biophysical models, using airborne hyperspectral (430-12700 nm) data, and assess the transferability of the approaches proposed to imagery of lower spatial and/or spectral resolution acquired by current and future spaceborne multispectral and hyperspectral sensors (EnMAP, HyspIRI, Sentinel2). Emphasis in this part of the research will be on integrating methods for increasing endmember separability and for reducing endmember variability in the unmixing. Here we will also take a closer look at the issue of thermal unmixing and the use of LiDAR height profiles in spectral unmixing. - Examine the use of spectral, thermal and LiDAR remote sensing data for characterising chemical and structural properties of urban vegetation. Again we want to assess the transferability of the approaches proposed to imagery of lower spatial and/or spectral resolution acquired by current and future spaceborne sensors. Emphasis in this part of the research will be on developing and validating generic spectral and structural indicators that allow for a proper quantification of the chemical (e.g., chlorophyll, water) and structural (e.g. LAI, biomass) properties of urban vegetation. Specific attention will be given to the development of novel signal unmixing approaches that allow removing, or adjusting/correcting for, the background spectral contributions from a mixed reflectance signature and thereby potentially improve the accuracy of existing vegetation indices. These objectives will lead to generic image processing tools for characterising urban land cover, including a functional characterisation of the urban vegetation derived from different sensor platforms. At the same time our analysis will provide improved insight in which level of detail (which types of materials, vegetation types, characteristics) can be provided at different spectral and spatial resolutions as an input for urban biophysical modelling (urban heat, hydrology).
Season: Preferred and acceptable dates: between June 2015-September 2015
Weather constraints: The proposed remote sensing activities require clear sky conditions, although some Cumulus clouds can be accepted if not positioned on the crucial target area. Windless conditions are preferable, as high wind velocities may influence the heat transfer and as such disturb the measurements in the thermal part of the spectrum.
Time constraints: The flight campaign is preferably organized during the summer season. This choice is mainly motivated by: 1) The vegetation growing season. During summer period (at the height of the growing season) functional differences/properties (e.g. LAI) are most distinctive. 2) As stated by Kutsch et al. (2001), surface temperature data are very sensitive to "abiotic noise", mainly wind and sudden changes in air temperature and cloud cover. Therefore, cloudless and windless summer days with unlimited incoming sun-radiation are to be preferred for taking surface temperature measurements (Wagendorp et al., 2006). Kutsch et al. (2001). Environmental indication: a field test of an ecosystem approach to quantify biological self-organization. Ecosystems, 4, 49-66. Wagendorp et al. (2006). Land use impact evaluation in life cycle assessment based on ecosystem thermodynamics. Energy, 31, 112-125.
Flights (number and patterns): We anticipate 1 flight per study area (see previous section for detailed locations). So, three acquisitions in total. If it would appear that this scheme is not completely feasible within the time constraints a downscaling of the flight campaign can of course be discussed. In that case priority should go to the Brussels study area. Due to the high level of heterogeneity in urban ecosystems, a high spatial resolution (2.5 to 10m) is required. As a result the flight height will vary between 975m and 4000m (see also section "Other constraints or requirements"). The number of flight lines per study area will of course depend on the flying height. Details on the study sites to be covered are provided in the previous section. The orientation of the flight lines should preferably be as much as possible into the solar plane (+/- N-S, S-N flight pattern) for homogeneous illumination.
Instruments: None
Other constraints: For the Brussels study site: due to the vicinity of the airport in Zaventem (in the north) and the presence of specific landmarks (e.g. Royal Palace), there might be some restrictions in defining the flight plan (height, orientation). However, a flight should not be a problem as in previous years an airborne CASI campaign (2002) and APEX campaign (2011, 2013) took place for the same study area. However, the issue should be taken into account during the flight planning.
Name: SOMERS Ben
PI email: ben.somers@ees.kuleuven.be