ALGASED - Remote sensing for characterization of intertidal sediments and microphytobenthic algae

Context and objectives

The sites studied in ALGASED are intertidal areas containing zones of highly dynamic, soft, muddy sediments and are of high ecological and/or economical importance. Monitoring and understanding the important processes of these intertidal flats are of interest for decision makers. Erosion and deposition of these sediments are mainly influenced by the sedimentological, biological, and physical factors. Therefore, remote sensing data , along with field data, were used to characterize the main intertidal sediment habitat types .

The main objectives of ALGASED were regarding several topics:

  • Modeling primary production
  • Attaining MPB biomass and its influence on the physical properties of sediments and indices
  • Assessing the performance of satellite data for the quantification of MPB and sediment physical properties
  • Advice on band selection for future missions
  • Spatially studying sediment characterization
  • Assessing and improving supervised and unsupervised classification techniques for sediment characterization
  • Creating a database that integrates the available data and metadata on the study sites

Project outcome

  • Work in ALGASED has been presented in international workshops.
  • Building simulated datasets of Hyperspectral data based on sediment biophysical properties
  • Spectral resampling of hyperspectral and ASD data into multispectral data
  • Semivariogram analysis and modeling – directional and omnidirectional
  • A comparison between the results of k-means, Gustafson -Kessel, and mixture of Gaussians clustering techniques on simulated data : this includes an automatic labeling of clusters, confusion matrix, and kappa statistic
  • Algorithms for estimating the amount of MPB using several existing and new indices
  • A multi scale model of the Molenplaat (PCNM analysis)
  • Guidelines about the usefulness of satellite data to study intertidal sediment and ecosystems related to the spectral and spatial resolution
  • List of the main variables determining estuarine ecosystems (with positive and negative correlations and feedbacks between the parameters)
  • Description of the influence of moisture content and mud fraction on the indices and ratios that qua ntify chlorophyll a absorption
  • Guidelines about the usefulness of satellite data to study intertidal sediment and ecosystems related to the spectral and spatial resolution
  • Guidelines of building hyperspectral simulated datasets based on biophysical sediment properties
  • Guidelines of unsupervised classification – choosing the number of clusters, features, and technique
  • Guidelines regarding the use of unsupervised classification to choose sediment property classes for supervised classification
  • Guidelines to carrying out a geostatistical analysis to study spatial structures on an intertidal flat
  • Maps of the primary production on the Molenplaat based on biomass estimates and emersion time
  • Maps of the main sediment types for the study areas and for different years using unsupervised and supervised classification methods
  • An extensive database was created of all the in situ measurements, hyperspectral and multispectral imagery, and metadata.