PSCIS

moved to https://github.com/smnorris/bcfishpass/tree/main/01_prep/02_pcis

The BC Provincial Stream Crossing Information System (PSCIS) monitors fish passage on road-stream crossings throughout British Columbia. This repository downloads PSCIS data to a Postgres db and links the crossings to the BC Freshwater Atlas - permitting upstream/downstream queries that enable reporting on linear habitat potentially blocked by failed culverts.

Requirements

a PostgreSQL/PostGIS database (tested with 12.1/3.0.1)
bcdata
BC Freshwater Atlas data loaded and configured via fwapg
BC Fish Passage potential habitat model (available on request from the Fish Passage Technical Working Group):
- fish_passage.modelled_crossings_closed_bottom
- fish_passage.modelled_crossings_open_bottom

Installation / Setup

Download the scripts:

$ git clone https://github.com/smnorris/pscis.git

Once the FWA database is set up via fwapg no further setup should be needed. The scripts assume that your database connection paramaters are stored as environment variables ($PGHOST, $PGUSER etc).

Data load

Load the PSCIS tables to the whse_fish schema, plus the lookup matching PSCIS points to FWA streams and modelled crossings (data/pscis_stream_matching.csv).

./01_download.sh

Load the modelled crossing tables (closed bottom/culverts and open bottom/bridges) to the fish_passage schema. For example:

psql -c "CREATE SCHEMA IF NOT EXISTS fish_passage"

ogr2ogr \
  -f PostgreSQL \
  PG:host=localhost port=5432 user=postgres dbname=postgis password=postgres \
  -lco SCHEMA=fish_passage \
  -lco GEOMETRY_NAME=geom \
  fish_passage_points.gpkg \
  modelled_crossings_closed_bottom

ogr2ogr \
  -f PostgreSQL \
  PG:host=localhost port=5432 user=postgres dbname=postgis password=postgres \
  -lco SCHEMA=fish_passage \
  -lco GEOMETRY_NAME=geom \
  fish_passage_points.gpkg \
  modelled_crossings_open_bottom

Usage

Run the sql scripts in order:

./02_clean.sh

Method

The sql scripts / output tables are:

`01_pscis_points_all`

Combines the four WHSE_FISH.PSCIS views (assessments, confirmations, designs, remediations) into a single table with unique crossings.

`02_pscis_events_prelim1`

Reference pscis_points_all to the closest stream(s) in FWA stream network within 100m. A crossing will often be within 100m of more than one stream, all results are kept in this preliminary step.

`03_pscis_events_prelim2`

From events_prelim1, attempt to find the best matching of the pscis point to stream using a combination of:

distance of PSCIS point to stream
similarity of PSCIS stream_name column to the gnis_name of the stream
the relationship of PSCIS downstream_channel_width to the stream_order of the FWA stream - if a very wide channel is matched to a low order stream, it is probably not the correct match

`04_modelled_crossings_all`

Combine all modelled crossings (fish_passage.modelled_crossings_closed_bottom, fish_passage.modelled_crossings_open_bottom) into a single table containing all modelled road crossings.

`05_pscis_model_match_pts`

Match PSCIS points to the modelled crossings where possible. Similar to the matching of PSCIS crossings to streams noted above, the script attempts to find the best match based on:

distance of PSCIS point to modelled crossing point
matching the crossing types (if PSCIS crossing_subtype_code indicates the crossing is a bridge and the model predicts a bridge, the points are probably a match.
as above, check the relationship of PSCIS downstream_channel_width to the stream_order of the FWA stream

`06_pscis_events_prelim3`

Combine the results from 1 and 2 above into a single table that is our best guess of which stream the PSCIS crossing should be associdated with. Because we do not want to overly shift the field GPS coordinates in PSCIS, we are very conservative with matching to modelled points and will primarily snap to the closest point on the stream rather than a modelled crossing farther away.

`07_pscis_events`

Create primary output table of interest - all pscis records (that are on a stream) as points on the FWA stream network. Do this by first adding all PSCIS records that have been manually matched to streams to the event table (currently BULK, ELKR, HORS, LNIC). Then for remaining points, remove locations from pscis_events_prelim3 which are obvious duplicates (instream position is within 5m). The PSCIS feature retained is based on (in order of priority): - status (1 REMEDIATED, 2 DESIGN, 3 CONFIRMATION, 4 ASSESSED) - most recently assessed - closest source point to stream

`08_pscis_points_duplicates`

For general QA of the PSCIS database, create a report of all source crossing locations that are within 10m of another crossing location.

`09_pscis_model_combined`

DRAFT - a first rough cut at combining the PSCIS data and modelled road-stream crossings into a single table for easy upstream / downstream barrier analysis.

To go further with this, we need to determine how to structure and work with our barrier analysis database:

include non-barrier PSCIS structures
include modelled OBS?
include barriers other than road crossings? (Dams, natural barriers etc)

`10_cleanup`

Clean up, deleting temporary tables created by above steps.