Note : this file contains many plots; it might take a few seconds to fully load.
Note2 : the interactive plots require an internet connection for full functionality.
Radar data on nocturnal bird fluxes were collected by horizontal radars at Luchterduinen and Borssele Alpha (Figure 2.1). The radar at Luchterduinen has been operational since spring 2019, and at Borssele since autumn 2019. Radar data is filtered before calculating fluxes (also called mean traffic rates (MTR)). These processing filtering steps are required to distinguish between tracks of birds from noise (e.g. rain, waves, slow moving objects, reflection by turbines). Filtering and MTR calculation were carried out by RWS using methodology developed by the UvA for the horizontal radar (Bradarić 2022). For more information on the filtering and calculation of MTR we refer to Kraal et al. (2024).
The bird migration model was developed by Bradarić (2022). The purpose of this model is to predict bird flux at Luchterduinen based on predictions of meteorological parameters. Massive bird migration occurs in autumn and spring. Since migration patterns differ between these two seasons, two modules were developed. The autumn module of the current version (1.1) of the operational model was trained on radar data from October and November from the period 2019-2023, and the spring module on data from the period 15 February – end of April over the period 2019-2023 (pers. comm. M. Braderic).
The interactive figures below show the cumulative percentage of total nocturnal migration per season per year for which horizontal radar data was available. All fluxes of available hours, measured by the horizontal radar, were ranked twice in descending order. First they are ranked based on the measured MTR by the horizontal radar (hr). Second, the same fluxes by the radar are ranked by the modeled MTR’s by the UvA model (mod). Next, cumulative MTR’s are calculated and translated to a percentage of the total migration measured by the radar for that season/period. These hours are plotted in the figures below, where the hours with the highest measured MTR’s are on the left side of the figure).
Cumulative plots are created for both offshore wind farms; Luchterduinen (=LUD) and Borssele Alpha (=BSA), and for each year/season where data were available, and for different lengths of the season; Full season (=3.5 months) and Short season (=only the months for which the model was developed). Cumulative plots are also created per offshore wind farm per season for all years together.
It is IMPORTANT to realize that much of the data was also used for training of the model. The match between measured MTR by the radar and modeled MTR in years used for training will likely be better than for a future season.
These graphs can be used to explore the number of hours (N curtailment hours, x-axis) needed for shutdowns in order to allow a safe passage for a certain percentage of seasonal migrants (Cumulative measured MTR (%), y-axis) and what the threshold would need to be if the model perfectly represents the radar measurements (red line) and with the current model performance (teal line). All this information pops up, when you hover over the lines with your mouse. The threshold based on model data is also known as the triggervalue.
Bradarić, M., 2022. On the radar: Weather, bird migration and aeroconservation over the North Sea. PhD Thesis. University of Amsterdam, Amsterdam.
Kraal, J.J., R.P. Middelveld, R.S.A. van Bemmelen, A. Gyimesi 2024. Validation of the outcomes of the bird migration prediction model for spring 2024. Report 24-303. Waardenburg Ecology, Culemborg