Context and objectives
In Spanish and Irish waters jellyfish blooms have caused considerable damages to tourism and aquaculture, respectively. The JELLYFOR project aimed to develop a medium-term forecasting model to mitigate the effects of these blooms. Because of the lack of life cycle models, a numerical model was created for the prediction of jellyfish. The model was trained with in situ measurements (temperature, salinity, chlorophyll concentration) and jellyfish observations, coupled with satellite-derived ocean colour data (chlorophyll and suspended matter concentrations and temperature). A near-real time service was required for locations which have sufficient jellyfish observations coupled with measurements.
Objectives for the Belgian part of the project (JELLYFOR-BE) were the reporting of the state of the art, the set up of an archive of satellite data for model training and a near-realtime service of satellite data to the model, and quality control and validation of the satellite data. An assessment of the observation and forecasting of jellyfish in Belgian waters was made during the last part of the project.
A highly automated and generic system for processing ocean colour data from MODIS and MERIS was developed and used to provide historical and near real time input for the jellyfish forecasting model running in the Spanish/Catalan and Irish regions
The drift modelling approach (Duliere et al. 2014) was tested for two jellyfish swarm beaching events that occurred along the Belgian coast in 2013 and provided information on the probable origin of those swarms. This model approach is considered to much more robust than the non-linear pattern recognition approach, but for forecasting requires input on the current location of any jellyfish swarm.
The high resolution satellite data from Landsat-8 seems to be insufficient to detect jellyfish swarms. The extremely high resolution data from Pléiades should theoretically be sufficient for detection of the larger individuals. However in the Pléiades data processed so far for Belgian waters, there are no obvious jellyfish swarms, although it is possible that with a better understanding of the spectral properties of jellyfish this might become feasible already with Pléiades.
In an important and unforeseen spin-off of the processing of Landsat-8 imagery in Belgian waters, turbid sediment wakes were discovered associated with offshore wind farms.
The feasibility of the Jellyspec camera/spectroradiometer system has been demonstrated for the measurement of jellyfish optical properties.
|Project leader(s):||IRSNB/KBIN - Directorate Natural Environment - Ecosystems data processing and modelling|