search

jncc / bh3-automation

Automation of the Extent of Physical Damage marine biodiversity indicator
MIT License
1 stars 1 forks source link

Categorizing OSPAR SAR values #5

Open HartmutT opened 5 years ago

HartmutT commented 5 years ago

The OSPAR layers have sar_surface and sar_subsurface columns, whose values range from 0 to 70.39 and 13.0, respectively. The expressions for computing SAR scores are applied to a column named SAratio whose values do not exceed 3. It appears that the SAR categories expressions are not applicable to the values in those two columns, but these are the only SAR columns available. How are their values to be categorized?

LRPettit commented 5 years ago

I think that the “SAratio” name in the expression is just a generic label and has to be modified according with the name of the column containing abrasion information in the pressure layer. Pressure layers (i.e. “ICES.2018.OSPAR-spatial-data-fishing-intensity” and “ICES Abrasion Layers 2009 to 2016”) report abrasion levels as swept area ratios and these need to be converted into categorical scores ranging 0-5. The expression for computing SAR scores says that if the swept area ratio values is higher than 3, than the corresponding categorical values will be 5. Please note that the columns containing swept area ratios values have different name in the two pressure layers: In the layer “ICES.2018.OSPAR-spatial-data-fishing-intensity” info are in the "SubsurfSAR" and "SurfSAR" columns. In the layer “ICES Abrasion Layers 2009 to 2016” info are in the "SurfaceSAR" and "Subsurface" columns.

HartmutT commented 5 years ago

I'm not worried about the field names, but the expressions in the "cookbook" suggested a different measurement scale, dividing the range 0-5 into roughly equal segments, whereas the values in the OSPAR layers range from 0 to 70. I had not imported the fishing density shape files. There are multiple, partially overlapping files per year. How are these overlaps to be handled? Do we use the max surface and subsurface abrasion value from spatially equal c-squares?

HartmutT commented 5 years ago

Another question about the fishing density files. Are these supposed to be an alternative source of pressure data or are they to be combined with the ICES Abrasion layers? If the latter, how are values from overlying squares to be combined (max of all?)?

HartmutT commented 5 years ago

Having taken a close look at the fishing density files, I gather that the "total_" SHP files are the same (?) as the ICES Abrasion Layers 2009 to 2016.gdb and the other SHP files are the individual data sources aggregated in the "total" files. Is this correct? Does this mean we don't need the SHP files?

LRPettit commented 5 years ago

I’m not sure that this answers your question, but If SAR is above 3 (it doesn’t matter how much above) then the categorical score will be 5. It doesn’t need to be split into equal segments.

The fishing density files are alternative sources of pressure data and not combined. Their geographical range is different and they cover slightly different time periods.

The SHP files in the ICES.2018.OSPAR folder are split by different gears and are also amalgamated into a total. These total SHP files should be the same as the total in the ICES Abrasion Layers, but the OSPAR covers a greater geographical area and a slightly longer time period. The SHP files for the individual gears are not always needed and the total can be used, but sometimes, we may want to just look at one particular fishing gear type.