This repository includes code for estimating pollutant loads to Norwegian coastal waters. The work is part of a joint monitoring programme under the OSPAR Commission for the Protection of the Marine Environment of the Northeast Atlantic. The main purpose is to estimate total loads of selected pollutants draining annually to Convention waters from inland water bodies.
The Norwegian component of the programme involves collaboration between NIVA, NIBIO and NVE, and is supported by the Norwegain Environment Agency. The code available in this repository was produced by NIVA.
The links below provide background to the project and develop code for the new project workflow.
-
Initial data exploration. Getting to know the project and exploring the results from previous years
-
Updating discharge datasets. Adding new observed and modelled flow datasets from NVE
-
Estimating loads at "monitored locations". 155 sites in the Norwegian component of the programme have at least some water chemistry measurements. This notebook estimates pollutant loads for these locations, using a combination of observed flows & discharges, modelled flows and regression analysis. Results for 2015 are estimated using the new code and compared to those reported previously
-
Preparing input files for TEOTIL and running the model. For parts of Norway where monitoring data are unavailable, the model TEOTIL is used to estimate nutrient loads (nitrogen and phosphorus). This notebook describes the data processing required to generate input files for the model and compares the output from the new workflow with previous results.
-
Processing annual inputs for unmonitored datasets. Experiments with TEOTIL suggest the workflow for processing input datasets can be simplified. This notebook restructures the raw annual data and adds it to the database. The database can then be used to generate input files for TEOTIL or NOPE.
-
Summary tables in Microsoft Word. Preparing key data tables for the report
-
A new export-coefficient-based pollutant model for Norway. Some changes are required to the way loads are modelled in the RID project. One option is to modify TEOTIL, another is to develop a new model entirely. This notebook develops and tests a simple export-coefficient-based pollutant model (provisionally called NOPE) for simulating Norwegian river loads.
-
Estimating loads in unmonitored areas. Using the new NOPE model to estimate loads in unmonitored regions, and comparing the results to values previously reported.
-
NOPE autocalibration. Using Bayesian MCMC to calibrate and evaluate the new NOPE model.
-
NOPE Metals. Prototyping a new NOPE module for simulating metal fluxes.
Links to the annual data processing are provided below.
-
Data processing for "monitored" locations. The code developed above is applied using data for 2016 for the 155 sites where water chemistry is measured.
-
Submit raw water chemistry data to Vannmiljø. Each year, raw water chemistry data collected by the RID Programme are submitted to the Vannmiljø database. This notebook compares the Vannmiljø submission template generated by Aquamonitor for 2016/17 to the quality-assured values in the RESA2 database.
-
Updating OSPAR discharge summaries. Some revisions are required to discharge data previously submitted to OSPAR. This notebook applies the corrections and fills-in new OSPAR templates for the period from 1990 to 2016.
-
Processing annual inputs for 2016. Restructuring the land use, sewage, industry and fish farm datasets for 2016 and adding them to the database.
-
Estimation of pollutant loads for unmonitored areas in 2016. Using NOPE to estimate total N and P in unmonitored regions for 2016, and summing point discharges for metals.
-
Generate overall summary table. Combine monitored and modelled outputs to generate a set of tables summarising national scale pollutant loads.
-
Fill-in OSPAR reporting template. Summary results from the from the project as a whole are aggregated and reported to OSPAR in a single Excel template.
-
Changes to the programme for 2017 to 2020. The monitoring programme for 2017 to 2020 is different to that followed previously. This notebook describes the main changes and implications for subsequent analyses.
-
Data processing for "monitored" locations. Empirical load estimation using 2017 data for the 155 water chemistry stations.
-
Count Limit of Detection (LOD) values. Estimate the proportion of LOD values for each parameter during 2017.
-
Average monthly water temperatures for 2017. Combining various temperature datasets to estimate mean monthly temperatures for the 20 main rivers.
-
Processing annual inputs for 2017. Restructuring the land use, sewage, industry and fish farm datasets for 2017 and adding them to the database.
-
Estimation of pollutant loads for unmonitored areas in 2017. Using NOPE to estimate total N and P in unmonitored regions for 2017, and summing point discharges for metals.
-
Fill-in OSPAR reporting template. Summary results from the from the project as a whole are aggregated and reported to OSPAR in a single Excel template.
-
Data processing for "monitored" locations. Empirical load estimation using 2018 data for the 155 water chemistry stations.
-
Count Limit of Detection (LOD) values. Estimate the proportion of LOD values for each parameter during 2018.
-
Average monthly water temperatures for 2018. Combining various temperature datasets to estimate mean monthly temperatures for the 20 main rivers.
-
Processing annual inputs for 2018. Restructuring the land use, sewage, industry and fish farm datasets for 2018 and adding them to the database.
-
Estimation of pollutant loads for unmonitored areas in 2018. Using NOPE to estimate total N and P in unmonitored regions for 2018, and summing point discharges for metals.
-
Fill-in OSPAR reporting template. Summary results from the from the project as a whole are aggregated and reported to OSPAR in a single Excel template.