Marine ecosystems are defined not only by the general abundance of primary producers, but also by their variability. In fact, biological production in many regions of the ocean is punctuated by hotspots and blooms that exhibit a high spatial and temporal variability in the phytoplankton biomass. A classic example of how periodic changes in primary production have profound effects on the food web and export of carbon is the annual North Atlantic spring bloom.
Satellite-derived proxies for phytoplankton biomass such as Chlorophyll (Chl) and Particulate Organic Carbon (POC) provide unprecedented coverage in time and space to better estimate the variability in phytoplankton biomass on different scales. However, one key challenge when working with satellite-derived data arises from data gaps due to masking of the ocean by clouds and sunlight reflected off the surface. In fact, on average, only 20% of the derived Chl fields prove useful. The erroneous data are also not evenly distributed but show a patchiness that reflects the temporal and spatial scales of synoptic regional weather systems. Such sparse and unevenly distributed datasets create a major challenge for common time-series analysis tools, such as Fourier analysis or Empirical Orthogonal Functions (EOFs), thus hindering efforts to understand the frequency distribution of the data.
True color image of the ocean covered by clouds. Photo credits: NOAA.
To better meet the specific challenges with time series analysis of sparse satellite-derived properties, PRIMUS members Bror Jönsson and Shubha Sathyendranath from PML have developed a new method to estimate dominating timescales of variability from datasets where up to 90% of the data is missing: MOving Standard deviation Saturation (MOSS). The paper has just been published in Remote Sensing of Environment and represents the first scientific output from PRIMUS.
The MOSS approach is similar to semi-variograms and earlier analyses of spatial patchiness but describes temporal variability rather than the spatial autocorrelation or patchiness in the satellite field. The technique is based on calculating the standard deviation (σ) of the time series data over moving windows of a set time interval and repeating for different time-interval windows. The average σ for each window size () increases from zero for a time window that includes just one data point, to σ when considering the entire time series. The largest possible time window is, in effect, the full time series. The shape of the resulting curve of vs. the window size (w) is then analyzed to identify a dominating time scale, τd of the time series based on the half saturation constant (KM).
The image above depicts the dominant timescales () obtained from MOSS, of (A) Sea Surface Temperature from the ESA Climate Change Initiative (SST-CCI) and (B) Ocean Colour Climate Change Initiative (OC-CCI,) for the period 2006-01-01 to 2016-12-31. Regions with a temporal coverage of less than 10 % are shaded.
Results show that the method can assess dominating timescales in time series where data coverage is sparse. They find that τd values for Chl and SST are not consistent or correlated with each other over large areas, and that, in general, SST varies on longer timescales than Chl, i.e. τd (SST) > τd (Chl). There is a threefold variability in τd for SST and Chl even within regions that are traditionally considered to be biogeographically homogeneous. The largest τd for Chl is generally found on the equatorial side of the trade wind belts, whereas the smallest τd is found in the tropical Pacific and near coasts, especially where upwelling is common, such as in Atlantic EBUS systems studied in PRIMUS. If the temporal variability in Chl and SST were driven largely by ocean dynamics or advection, regional patterns of τd for SST and Chl should co-vary, which can be seen in coastal upwelling zones. The lack of coherence between τd (Chl) and τd (SST) in other regions suggests that biological processes such as phytoplankton growth and loss decouple the timescales of Chl variability from those of SST and generate shorter-term variability in Chl.
Their findings are novel, but they also have the potential to help elucidate patterns in observed global phytoplankton biodiversity. For instance, small organisms might have a competitive advantage in regions with short dominant timescales. Another potential use could be to constrain estimates of carbon export from the surface ocean by identifying regions where timescales of biomass variability may play an important role. The authors believe that the MOSS method can provide a framework for a concerted approach to connect variability in phytoplankton biomass and biological production with export production and export efficiency.
If you want to know more about this novel study, click here.
Jönsson, B.F., Salisbury, J., Atwood, E.C., Sathyendranath, S., Mahadevan, A., 2023. Dominant Timescales of Variability in Global Satellite Chl and SST revealed with a MOving Standard deviation Saturation (MOSS) approach. Remote Sensing of Environment 286, 113404.