Dataset for Integrated Species Distribution Model for pikeperch larvae in the Porvoo-Sipoo archipelago

This record contains the data required to run the code for fitting the Integrated Species Distribution Model described in arXiv:2206.08817 [stat.ME].

Files in this record

transect_data.csv Line transect observations from Porvoo-Sipoo archipelago, Finland on June 2017.

expert_assessments.tif Rasterized, anonymous expert assessments. Categorical values denoting how likely a given location is to be a spawning location for pikeperch. 4 categories, with smaller values corresponding to higher probabilities.

covariate_raster_example.tif Rasterized example environmental covariate values. These are similarly structured as the covariate data used in the study and compatible with the analysis code. However, since we do not have the permission to release the original data set, these values are instead generated based on the projected planar coordinates such that they have roughly similar spatial gradients as the original covariates.

Detailed descriptions

Transect data

Location and replicate identifiers

id : transect identifier. Replicates of the same transect have the same identifier.

id2 : alternate transect identifier, unique for each transect.

repeated : whether transect was replicated or not.

X_euref : easting coordinate, EUREF_FIN_TM35FIN, for the transect starting location in [meters]

Y_euref : northing coordinate, EUREF_FIN_TM35FIN, for the transect starting location in [meters]

date : date of the measurement, DD/MM/YYYY

week : week number of the measurement date

In situ measurements

volume : Transect water volume [m^3]. Transect length (500m) multiplied by sampler surface area. Used as survey effort.

heading : compass heading (direction) for the transect, in [degrees].

SumKUHA : total pikeperch (Sander lucioperca, kuha in Finnish) larvae count in each transect [scalar]

Expert assessments

The raster contains assessments from 10 local experts encoded as separate raster layers (Expert_1, Expert_2, ..., Expert_10). Raster resolution is 50m x 50m and the planar coordinates are based on the same coordinate reference system as the transect observations (UTM zone 35).

The assessments are coded as integers with values between 1 and 4, with smaller values corresponding to higher probabilities.

Covariate raster example

This raster has the same spatial dimensions and uses the same coordinate reference system as the expert assessment raster and has three layers, one for each covariate. The covariate values are generated based on the spatial coordinates such that each covariate has similar spatial gradient as the original covariate. The covarites have the same names as in the original covariate data (dptLUKE, dist10m and lined3km).

Creators

Transect data collected and curated by Sanna Kuningas.

Original covariate rasters curated by Sanna Kuningas from data sets collected by the Finnish Environment Institute and the Natural Resources Institute Finland.

Expert assessments originally digitized and rasterized by Jussi Mäkinen.  Additional refinement to assessment rasters by Karel Kaurila.

Preparation for publishing on Zenodo for all of the data sets  by Karel Kaurila.

Change log

 2025 Jan 31: Included columns date and week for transect_data.csv.