Satellite Data Visualization

Option 1:
Load a file with predefined settings (.txt, created in PhytoFit): Browse to select, then click "Apply settings" and "Load data" below.


Option 2:
Start selecting your settings below, then click "Load data" and adjust remaining settings as needed.


Region
Satellite and variable
View full satellite [chla], or use one of two models to separate satellite [chla] into concentrations of different phytoplankton cell sizes, and choose the cell size to view:

Year
Data composite length

Log map color scale and axes on plots? (This does NOT log the actual data)


(Optional) Enter a name (use only alphanumeric characters, underscores, or periods) and click "Apply".
Choose the method to create your polygon.
Draw polygon using the toolbar at the top left corner of the map.
Enter decimal latitudes and longitudes for vertices of polygon, separated by commas, then click "Create polygon". Use lon/lat < 0 for west/south. Lists must be the same length, with >2 values each, in the same order so that each latitude is paired with longitude.
Click "Browse" to find a shapefile. Select the "shp" file and all files with the same name but different extensions (e.g. dbf, prj, sbx...), then "Open". Shapefile must contain a Simple Features (sf) object. If the sf contains multiple polygons, a button will appear below to select the polygon you want to use. WARNING: polygons with a large number of vertices may take several seconds to load.
List of latitudes:
List of longitudes:



Minimum % coverage.
Composites with lower coverage in the selected polygon will not be plotted on the density plot or time series, or used in the model fit.
Outlying pixel detection method
Remove outlying pixels before calculating statistics for the selected composite and polygon.
SD = standard deviation
IQR = interquartile range
Statistic
Calculate either the mean or median value of each composite, to be used in the time series and model fit.
Range of pixel values
Choose the range of values allowed in the calculation of the statistics and model fit (pixels outside this range will be omitted).
If a limit is left blank, it will be ignored.
–


Fit a model to the points on the time series.
Choose fit method, model shape, and point smoothing method.
LOESS span
Controls the degree of smoothing.
Fit the model to log10-transformed data?
Set the range of days to use in the model fit, and the limits of the day of peak amplitude and the first day of the bloom.
Switch to ON to weight each point in the time series by percent coverage.
When calculating magnitude (area under the curve between tstart and tend) and amplitude of the curve, should the background be removed first?
Flags
Fits will be flagged if they meet certain criteria that indicate potential problems with the fit (NOTE: this does not affect the fit itself). Combinations of flags will be written as a single number (for example, 13 for flags 1 and 3). Click below for details. Optionally adjust the parameters of some flags.
Select the method used to calculate tstart :
Either a percentage of the curve amplitude between 0.01 and 90 (peak minus background), or a constant threshold between 0.01 and 5 (difference between the fitted curve and background value, calculated in linear space).
Set % curve amplitude to mark start of bloom.
Set increase over background to mark start of bloom.
Switch to ON to consider tmax_fit a parameter in the regression.
WARNING: This will disable tstart limits.
Switch to ON to allow background values to vary linearly as a function of day.


Flag 1: Amplitude ratio limits
–
Flag 2: Magnitude ratio limits
–
Threshold coefficient
The start of the phytoplankton bloom is considered to be the point before tmax when [chla] drops below a threshold for > 14 days.
Threshold = chlamedian * threshold coefficient
chlamedian is the median of days with sufficient % coverage within trange.
NOTE: If you opted to log the input data, chlamedian is calculated by taking the median of the logged data and transforming it back to linear space to compute the threshold. The threshold itself is then logged again to determine the bloom initiation.



Time series
Select a series of years and the polygons you would like to process using the current settings, then click "Run time series" to generate the following:
  • tables of statistics (.csv),
  • a single .csv file containing the fitted parameters for all selected years and polygons, and
  • a .csv file containing the settings used for the time series.
Files will be zipped to a folder following the naming convention satellite_ region_ compositeLength_ years_ cellSizes_ variable_ fitmethod_ timecreated.
Make sure at least one polygon is selected.
When processing is complete and the new filename appears over the download button, click "Download results (.zip)".

Download results (.zip)

Save settings (.txt)


Download density plot (.png)

Download time series plot (.png) Download time series table of statistics (.csv) Download fit parameters (.csv)