How Numerical Weather Prediction NWP models help you assess air quality.

It's not quite this strict in Alberta

There are certain criteria for valid on-site meteorology: at least one year of hourly data, statistically related to longer-term weather, and at least 90% data availability. Missing data can be interpolated, corrected with established profiles, or supplemented with off-site data.

Other applications require five years of modelling using meteorological data from the Alberta Environment and Parks (AEPA) dataset, plus ground-based meteorology when available. On-site data should be handled the same way as missing data. AEPA data alone should be used if ground-based data isn't available or incomplete. Other meteorological datasets aren't acceptable.

Unusual place for a meteorological station

The Ministry of Environment has developed Regional Meteorological Data Sets for AERMOD, covering five years. AERMET and the Weather Research and Forecasting (WRF) model were used to prepare these sets. Modifications or alternate data sources may be needed if the facility has unique surface characteristics.

The dispersion of emissions can be stunted by calm wind conditions. For alternate meteorological stations, AERMOD skips hours with calm winds. There are a lot of ways to fill in gaps when there's missing data.

The different formats of meteorological data files are discussed, along with how to choose representative meteorological stations. Data accuracy depends on QA/QC procedures and adjustments for anemometer height and wind speed.

For accurate air dispersion modeling, selecting appropriate meteorological data and keeping its quality check are crucial.

Vancouver air quality forecast looks good

Different agencies provide meteorological data in BC, each with their own quirks. Among them are Meteorological Service of Canada (MSC), Ministry of Environment and Climate Change Strategy, which protects natural resources, and various fire weather networks. Siting, instrument type, maintenance, and data quality assurance are all things to consider when using diverse data. In BC, upper-air data are sparse and have to be carefully assessed for applicability prior to incorporating into the data set.

It's best to have up to five years of data for Level 2 or Level 3  Assessments, which are more detailed. Data completeness of 90% is recommended for use in AERMET and CALMET and it's a good idea to collect more data if this limit hasn't been met.

Air quality dispersion models rely on meteorological data as input, which can come from either predetermined conditions or actual observations. In the latter case, it's crucial to have quality-assured data to reduce uncertainty in model predictions.

For meteorological data, the BC guideline recommends site-specific collection programs, measurements from a similar location, or output from numerical weather prediction NWP models. To ensure data representativeness, a site-specific set of data is best. Different factors need to be considered, like geophysical similarity, climate regimes, instrumentation, and data quality, when considering alternate sources. 

It may be necessary to establish a new site-specific meteorological collection program for major projects in complex areas with no data sources.

There are several agencies in BC that provide meteorological data. Data like wind speed, direction, temperature, and cloud cover are included. If using these services, it's important to make sure the data is appropriate and compatible with the model.

Short gaps can be filled with interpolation techniques, while longer gaps can be filled with data from nearby sites or numerical weather prediction NWP models. It's important to acknowledge, however, that these approaches introduce uncertainty into the data.

AERMET and CALPUFF need specific data for dispersion modelling, including surface observations and upper-air observations. For long-term modeling assessments, data completeness and quality assurance are essential.

Air quality dispersion modelling relies on selecting appropriate meteorological data sources and making sure to use the best available data quality.

Using AERMOD/AERMET: Various meteorological data inputs are needed for AERMET, a tool for modelling air quality. In the absence of representative airport data, a site-specific meteorological program can be used, but cloud cover information is needed. AERMET can estimate cloud cover if necessary data isn't available.

British Columbia has complex terrain throughout the province, so getting the right upper air data can be tough. It's possible to use nearby soundings or NWP model output. AERMET also handles very low wind speeds, known as "calm" periods. Data format and requirements for integrating site-specific or airport data within AERMET are detailed, including wind speed, direction, temperature, and other parameters. SMERGE is a part of AERMET that helps prepare data files for AERMET. To model air quality accurately, AERMOD needs precise meteorological data formatted by AERMET and it must be based on good input information.

Graph the weather data

A separate file called PRECIP.DAT is needed for wet deposition calculations if applicable. This file can be prepared with preprocessors like PXEXTRACT and PMERGE, although Canadian agencies supply data in a different format. There are limitations to NWP model output, but it can supplement observational data or replace it.

NWP models can provide meteorological data for areas that lack complete observations, but quality and applicability must be considered carefully.  We have about different ways to generate meteorological data for dispersion modelling, which helps us understand how pollutants spread. Here's how it works:

• Numerical Weather Prediction NWP Models: These are computer models that predict weather patterns. They're useful but take a lot of power. Local areas and fine-scale terrain features aren't always accurate.
• A hybrid approach combines NWP output with CALMET to improve accuracy, especially in complex terrain areas.
• This model can generate meteorological data using NWP data alone, observational data alone, or a combination of both. There are a lot of factors to consider, like terrain complexity and data availability.
• Factors like model expertise, terrain complexity, and data availability affect the choice and effectiveness of these methods.
• In CALMET, there are several user-defined parameters and options that require expert judgment. Inclusion/exclusion of certain data points and weighting of observational data are affected by these parameters.
• There are recommendations for when to use each method and how to handle common problems. MMIF is a program that can help you process NWP output for dispersion modelling.

Here's some detailed guidance on how to generate meteorological data for dispersion modelling, taking terrain complexity, data availability, and model expertise into account.

How to set up and run CALMET, which is used to model atmospheric dispersion. Here are some points, but for completeness check the instructions in the BC Air Quality Dispersion Guideline, Section 6. Especially the tables.

Setting it up:

• Select options like vertical velocity, slope flows, and extrapolation of surface wind observations.
• Don't include terrain-forced vertical velocity in the initial guess.
• We explain how to adjust vertical velocity, compute slope flows, and extrapolate wind observations.
  

Usage of data:

• Use numerical weather prediction NWP models for output or coarse CALMET fields to use as initial guesses.
• For CALMET, define parameters related to surface meteorological data and upper air soundings.
• Establish biases and minimum distances between upper air stations and surface stations.

Interpolating wind fields:

• Specify the radius of influence and weighting of observations for wind field interpolation.
• We explain how to use varying radius of influence and barriers to interpolation.

Modules for diagnostics:

• There are options for surface temperature, lapse rate, and initial guess wind components.
• The parameters for using surface and upper air wind components are detailed.

There are a few more modules. If you're using Lake Breeze, set the parameters. Here's how to do quality assurance and quality control (QA/QC). Evaluation of the model:

• There are recommendations for plotting and analyzing wind vector fields, surface wind speeds, and wind roses.
• There are also suggestions for modelling approaches like CALPUFF for light wind conditions and stagnation.

This passage serves as a comprehensive guide for setting up and running CALMET for atmospheric dispersion modelling, including data preparation, model configuration, and results evaluation.