WP A3

Long term relevance and aquatic leverage

Dr. Hans Kautsky
Department of Systems Ecology
Stockholm University, SE-106 91 Stockholm, SWEDEN
Phone: +46 8 164244
Email: hassek@ecology.su.se

Sammanfattning
Arbetspaketet belyser vilken skala som behövs för att kunna tillämpa lasermetodiken i syfte att kartera och följa långtidsförändringar på de grunda, vegetationsklädda bottnarna. Härvid kommer upplösningen av metodiken vara av avgörande betydelse för arters identifiering och kvantifiering. Förmågan att kunna särskilja arter från varandra och vegetation från substrat, är av stor betydelse för att i framtiden kunna rationalisera t.ex. habitatkartering och miljöövervakning av stora områden. Därför är ett delmål att pröva ut ny teknik som kan särskilja arterna t ex via fluorescens.

Abstract
The WP A3 will address the issue of the resolution necessary to map and detect changes in vegetation covered substrates in shallow waters. The resolution from the laser technique is crucial for the detection of individual species and their quantitative distribution. Also, the ability to separate vegetation covered from bare substrate is essential. If we can detect species and their distribution, the laser technique can essentially improve habitat mapping and long term monitoring over large areas. It is essential to find new techniques, such as the proposed fluorescence method, to achieve the desired detail of species discrimination.

1    Purpose and Scope
The work package has two main purposes. The first aim will be to focus on long term use of the methods developed and evaluated in WP A1 and WP A2. Here, the methodological results will be checked for consistency over time, that is, a similar environment must give similar classification and mapping results when measured at different times. Additionally, the methods should also be adapted for changes and advances in laser and camera technologies so that today's mapping results could be compared to future mapping results with other and further developed technologies. In connection to this work, methods for change detection in the aquatic environment will also be developed. The second aim of this project will be to establish leverage with respect to possible technologies which would have increased potential for classification and mapping. Some examples are hyperspectral imaging, multispectral lasers and fluorescence remote sensing technologies. The priorities of the different efforts in WP A3 will depend on results from WP A1 and A2 and will be subject to the needs and suggestions from the Environmental Protection Agency.  

2    Contribution to the programme aim
WP A3 will specifically ensure that methods for mapping of aquatic habitats can be repeatedly performed for long term monitoring and change detection. One main problem, which will be addressed, is the scale of resolution needed for this purpose. Additionally, WP A3 will be open to new, emerging methods that are not fully operational today. These methods may have been tested in pilot studies, with prototype instruments, or with commercial equipment not yet fully operational in a broader sense. It should be noted that resolution can comprise both spatial, temporal and spectral resolution, which all have influence on the performance of the classification, mapping, monitoring and change detection.

3    Background, Theory and Methods
The priority between the activities will depend on the results obtained in WP A1 and WP A2 and on the requirements from the Environmental Protection Agency. The work will focus on prioritised areas among the following items.
    Adaptation and validation of successful and promising methods and user cases developed in WP A1 and WP A2. The methods will be validated for consistency over time (repeatability) and for ability to detect changes in the aquatic environment. These studies will require thorough field documentations of the test area with data collected in WP A2. One part of this activity should consider the effects of upgraded newer version of laser scanners. An example of this would be as follows: The aim is to quantify the increase/decrease of area covered by vegetation in a monitoring site.  Will improved resolution in the laser data allow comparison with data taken with lower resolution? Repeatability and change detection will be studied based on data from two separate occasions within the same geographical area, acquired in WP A1 (remote sensing data) and WP A2 (in-situ data), and processed by WP A1. 
    Improved classification of aquatic environments. Briefly stated, the ultimate goal for monitoring is to differentiate between vegetation and non-vegetated bottoms and also to differentiate between species. This concerns not only obvious cases such as the distribution of Zostera marina on soft substrates but also, for example, hard substrates covered with filamentous algae that have very subtle differences in colour and geometrical shape. In case the methods developed in WP A1 and WP A2 give poor classification and mapping results, new methods could be considered. One example is using laser techniques based on fluorescence, while another is using hyperspectral imaging data.
    Active colour images. Using laser illumination, active colour images can be produced by using a multicoloured laser and simultaneously monitoring the magnitude of the reflected light at several wavelengths. Images produced in this manner can reach larger depths than passive multi- or hyper-spectral systems (Jaffe 2001, 2005). Laser systems with registration of several wavelengths also allow for characterisation of inelastic, or trans-spectral, phenomena such as fluorescence. Fluorescence maps could possibly be produced that describe, on a point-by-point basis, the fluorescent characteristics of large and small individuals. Fluorescence is already used in water sample analysis (e.g., for quantification of chlorophyll content),  detection and imaging studies of vegetation (e.g., Edner et al. 1994), and in pilot studies for sea floor mapping from an underwater vehicle (e.g., Harsdorf et al. 1999). Today's laser scanners could, with slight modification, be used for fluorescence sensing of the sea floor, which could enhance classification capabilities. The ultimate goal is to detect the species composition and coverage based on the individual species fluorescence. In complex communities, the main constituents, referring to species covering more than 25 % of the substrate (e.g., Fucus vesiculosus, Pilayella, Furcellaria lumbrialis  etc.), may be separated from each other when the species specific fluorescence is known. Also, by separating signals from vegetation and from the substrate, the total coverage of the substrate could be detected. The change of species fluorescence with water depth is one main issue to be resolved for detecting the depth distribution of the species.
    Hyperspectral remote sensing systems, which obtain image data in many spectral bands, have been used in several studies for mapping of submerged aquatic vegetation. For example, Williams et al. (2003) studied classification of two species (Myriophyllum spicatum and Vallisneria americana) using HyMap imagery (126 spectral bands) with a pixel size of about 4 m * 4 m. Underwood et al. (2006) achieved classification of two aquatic invasives, Brazilian waterweed and water hyacinth, using the HyMap sensor (this time with a pixel size of about 3 m * 3 m). Their results indicate that classification of submerged aquatic vegetation is limited to fine spatial scales and that accuracy of mapping submerged vegetation across many flightlines is challenging owing to differences in water characteristics and tidal heights. Examples of work with hyperspectral remote sensing of the sea floor with the CASI-2 sensor are Tuell and Park (2004) and Phinn et al. (2008).
    Examples of activities within this WP are focused studies of classification possibilities with available state-of-the-art equipment within the project group  or with data provided by our external contacts. The processing of such data will be based on the correction methods (e.g., for depth and water turbidity) and tools developed in WP A1.

4    Practical Relevance
    The repeatability of the methods and their sensitivity in detecting changes are necessary to evaluate before the methods can be used for monitoring purposes. The scale of resolution, the ability to discriminate between main species and the depth limit for the methods will be addressed.  The research will further determine how new technologies can be applied to aquatic ecology by mapping of submerged aquatic vegetation in estuarine environments.  The work package will stress the requirements and environmental monitoring needs and give important information for future research.

5    References

  • Edner, H., Johansson, J., Svanberg, S., and Wallinder, E. 1994. Fluorescence lidar multicolour imaging of vegetation. Appl. Opt. 33:2471-2478.
  • Harsdorf, S., Janssen, M., Reuter, R., Toeneboen, S., Wachowicz, B., and Willkomm, R. 1999. Submarine lidar for seafloor inspection. Meas. Sci. Technol. 10:1178-1184.
  • Jaffe, J. S. 2005. Performance bounds on synchronous laser line scan systems. Opt. Express 13:738-748.
  • Jaffe, J. S., McLean, J., Strand, M. P., and Moore, K. D. 2001. Underwater optical imaging: status and prospects. Oceanography 13:66-76.
  • Phinn, S., Roelfsema, C., Dekker, A., Brando, V., and Anstee, J. 2008. Mapping seagrass species, cover and biomass in shallow waters: An assessment of satellite and airborne multispectral and airborne hyper-spectral imaging systems in Moreton Bay (Australia). Remote Sensing of Environment 112:3413-3425.
  • Tuell, G. and Park, J. Y. 2004. Use of SHOALS bottom reflectance images to constrain the inversion of a hyperspectral radiative transfer model. In Proceedings of Laser Radar and Technology Applications IX, Orlando, FL, 12–16 April 2004, SPIE Vol. 5412:185-193.
  • Underwood, E. C., Mulitsch, M. J., Greenberg, J. A., Whiting, M. L., Ustin, S. L., and Kefauver, S. C. 2006. Mapping invasive aquatic vegetation in the Sacramento-San Joaquin Delta using hyperspectral imagery. Environmental Monitoring and Assessment 121:47-64.
  • Williams, D. J., Rybicki, N. B., Lombana, A. V., O'Brien, T. M., and Gomez, R. B. 2003. Preliminary investigation of submerged aquatic vegetation mapping using hyperspectral remote sensing. Environmental Monitoring and Assessment 81:383-392.