Environment and Climate Change Canada showcases its research, development, and service initiatives in meteorology, air quality, and health during the Pan and Parapan American Games in 2015.
The Pan and Parapan American Games (PA15) were hosted in the Toronto, Canada, area from 10 to 26 July and 7 to 15 August 2015 and were the third-largest sporting event in the world. They include many sports that are considered for future inclusion at Olympic competitions. As a result, they pose a greater forecast and warning services challenge than the Olympic competitions as there are more outdoor venues for which to provide public severe weather, AQ (for a complete list of acronyms, please see the appendix), and health warnings.
Toronto, the fourth-largest city in North America, is located on the north shore of Lake Ontario (Fig. 1) with Lake Erie, Lake Huron, Georgian Bay, and Lake Simcoe lying to the southwest, west, northwest, and north, respectively (not shown). Lake-breeze (LB) circulations are frequently observed during the summer. Showers and thunderstorms often develop along the edge of LBs (Sills et al. 2002, 2011; Alexander 2012) and can result in severe weather such as tornadoes and flash floods (King et al. 2003; Boodoo et al. 2015). LB circulations also affect AQ as they reduce mixing, increase insolation, enhance secondary pollutant production, and transport pollutants and precursors both horizontally and vertically (Brook et al. 2011; Levy et al. 2010; Hayden et al. 2011). Major highways and roads transect the PA15 area and are a significant source of local air pollutants. LBs also affect temperature and humidity and can impact human health.
A map of the main PA15 area showing the location of the venues and organization of the mesonet. The labeled stations are vertical profiling sites. In the west is the Pearson supersite (YYZ), east is UOIT, north is the ECCC/DOW site, and south is HAN.
Citation: Bulletin of the American Meteorological Society 99, 5; 10.1175/BAMS-D-16-0162.1
A map of the main PA15 area showing the location of the venues and organization of the mesonet. The labeled stations are vertical profiling sites. In the west is the Pearson supersite (YYZ), east is UOIT, north is the ECCC/DOW site, and south is HAN.
Citation: Bulletin of the American Meteorological Society 99, 5; 10.1175/BAMS-D-16-0162.1
A map of the main PA15 area showing the location of the venues and organization of the mesonet. The labeled stations are vertical profiling sites. In the west is the Pearson supersite (YYZ), east is UOIT, north is the ECCC/DOW site, and south is HAN.
Citation: Bulletin of the American Meteorological Society 99, 5; 10.1175/BAMS-D-16-0162.1
ECCC was mandated to provide enhanced monitoring for the provision of venue-specific public hazardous weather, AQ, and health alerts and to support weather-sensitive emergency services, policing, and transportation. Unlike the Vancouver Winter Olympic Games (Doyle 2010; Joe et al. 2010), ECCC was not contracted to provide sport-specific weather predictions.
ECCC had very positive experiences with previous international forecast demonstration projects that focused on thunderstorm and winter complex terrain nowcasting (Keenan et al. 2003; Wilson et al. 2010; Duan et al. 2012; Isaac et al. 2014b; Kiktev et al. 2017). This included design and use of mesoscale networks (mesonet), the value of nested high-resolution models, advancements in nowcasting system development, and the challenge of technology transfer to operations and end users. A strategic insight was to exploit the prestige and firm deadlines of major events, such as PA15, to accelerate, coordinate, and demonstrate research and operational development projects. Within the context of PA15 and the related challenges of providing urban-scale weather, AQ, and health alerts and services, an integrated project called the Environment Canada Pan and Parapan American Science Showcase was initiated.
It initially focused on projects internal to ECCC and included testing and demonstration of new observation technology (surface, upper air, and radar), high-resolution meteorological and AQ modeling, point nowcasting, a prototype forecast production system, and the integration of new health initiatives. This paper presents an overview of the research and development activities. The operational and service activities are presented elsewhere [Johnston et al. 2017; ECCC 2016; more information can be found online in the supplemental information (https://doi.org/10.1175/BAMS-D-16-0162.2)].
The next sections provide a brief background, description of the observation and modeling systems, prototype forecast system, and the link to health services. The paper concludes with recommendations for future urban projects. Acronyms, observation system details, and meteorological overview are presented in the appendixes.
BACKGROUND.
Mesoscale features, such as LBs, urban heat island circulations, and air pollutant distributions, interact with each other, affecting severe weather, AQ, and heat episodes. Analysis of satellite and radar imagery showed that cumulus clouds develop along and at the intersection of low-level boundaries (Purdom 1976; Wilson et al. 1998; Wilson et al. 1994) that can also enhance existing and lead to the development of severe thunderstorms [tornadoes, severe wind gusts (>90 km h−1), large hail (>20 mm), and flash flooding]. The distribution and production rate of air pollutants are also affected by these boundaries (Brook et al. 2011; Levy et al. 2010). Contrasting temperature, humidity, and wind on either side of the LB can affect heat stress. Such boundaries include sea–LB fronts, thunderstorm gust fronts, drylines, and terrain-induced drainage and downslope flows.
In the Toronto area, the LB is a prominent cause of low-level boundaries (Wentworth et al. 2015; Sills et al. 2011). To extend the lead time for hazardous weather and refine AQ and heat alerts, a key science question is the characterization and prediction of LBs. The recent progress in high-resolution modeling has been remarkable, and key questions are at what resolution and with which parameterization schemes can salient features of the LB be captured and adequately predicted?
Heat is one of the primary causes of human mortality and morbidity (Prüss-Ustü;n et al. 2016). A dense urban mesoscale observation network was deployed for PA15, and black globe thermometers (BGTs) were added to initiate and support high-resolution health initiatives. This provided a leap in capability and the opportunity to demonstrate, for the first time, the impact and value of high-resolution (time and space) health predictions and alerts using both observations and numerical predictions to support public health response in risk-sensitive communities. Warning thresholds were based on health evidence and served to harmonize a patchwork of public health alerts, which had been in effect in the past and limited to observations made at airport-monitoring locations in the PA15 area.
THE OBSERVATION SYSTEMS.
The mesonet of surface stations was assembled from a variety of monitoring platforms and is summarized in Table 1, Fig. 1, and the appendix “Meteorological overview.”
Summary of observation systems. Status refers to whether the data were available to the ECCC data management and forecast systems before (existing), during (PA15), or after (legacy) PA15. Also see Fig. A1.
ECCC operated 30 existing automatic weather stations in the PA15 area. PA15 provided the opportunity to add three additional stations to fill critical gaps in the southwest Ontario area and the impetus to bring in data (70+ stations) from other agencies as part of a “network of networks” concept that remained after PA15 (see Table 1 for a list of agencies). Ultraviolet sensors were deployed at four locations in the PA15 area.
A total of 42 compact stations (27 WXT520 by Vaisala and 15 WS600/601 by Lufft) measured standard meteorological variables and were augmented by BGT used to calculate the wet-bulb globe temperature, a heat stress index (Fig. 2). The stations used solar panels and cell modems and were designed for autonomous operations and rapid deployment. The 1-min data were acquired (Isaac et al. 2014a), and derived variables (e.g., wind gusts) were defined on a rolling 60 min. An evaluation (summer only) demonstrated good comparison against references (Klaassen 2016; see supplemental information).
Three compact weather stations showing the typical nature of the site exposure. (left) WS601 sensor by Lufft is located at the Henley rowing venue. (middle) WXT520 sensor by Vaisala is located near the tennis venue at York University. (right) WS600 is mounted on the roof of the Toronto International Trap and Skeet venue.
Citation: Bulletin of the American Meteorological Society 99, 5; 10.1175/BAMS-D-16-0162.1
Three compact weather stations showing the typical nature of the site exposure. (left) WS601 sensor by Lufft is located at the Henley rowing venue. (middle) WXT520 sensor by Vaisala is located near the tennis venue at York University. (right) WS600 is mounted on the roof of the Toronto International Trap and Skeet venue.
Citation: Bulletin of the American Meteorological Society 99, 5; 10.1175/BAMS-D-16-0162.1
Three compact weather stations showing the typical nature of the site exposure. (left) WS601 sensor by Lufft is located at the Henley rowing venue. (middle) WXT520 sensor by Vaisala is located near the tennis venue at York University. (right) WS600 is mounted on the roof of the Toronto International Trap and Skeet venue.
Citation: Bulletin of the American Meteorological Society 99, 5; 10.1175/BAMS-D-16-0162.1
The mesonet was conceived to support the PA15 venue forecast and alerting programs as well as for high-resolution model validation. Hence, compact weather stations were located at the venues and situated to sample the LB along five transects (another line was approximately parallel to the shoreline) with higher spatial density near the shore. Data collection and quality assurance used the ECCC operational data management system. Siting weather stations in an urban environment was a challenge (Oke 2008), and 20 were mounted on one- and two-story buildings. Wind direction is a key parameter that is strongly affected by building effects and urban canyons. However, for the purposes of LB detection, the stations provided adequate observations.
ECCC operated three mobile weather units, AMMOS, over 22 days. Standard meteorological parameters, location, and particulate matter were sampled at 1-s intervals (Fig. 3). Western University (WU) operated another mobile weather station that included air and surface (road and wall) temperatures, humidity, and incident short- and longwave radiation (Wiechers and Voogt 2017). These mobile observations within the urban canyon focused on high-resolution simultaneous sampling (on the order of tens of meters) of the LB front where sampling by fixed stations was not possible. Sampling predominately occurred along north–south routes between the Lake Ontario shore and areas north of Toronto. In total, the vehicle traversed over 10,000 km, sampled 240 LB fronts and 7 thunderstorm outflow boundaries, along with 23 intra-urban traverses that sampled eight different neighborhoods.
(top left) A vehicle-mounted AMMOS unit and (top right) predetermined routes. Routes were mainly between the Lake Ontario shore and areas north of the highly urbanized parts of the greater Toronto area but could be extended west to the Milton area and north to Lake Simcoe as needed. The star marks the “home base” for the vehicles. (bottom) AMMOS data showing the rapid changes in temperature, dewpoint temperature, and relative humidity as the vehicle moved across the Lake Ontario lake-breeze front on the afternoon of 26 Jul 2015. Areas with blue shading are in lake air.
Citation: Bulletin of the American Meteorological Society 99, 5; 10.1175/BAMS-D-16-0162.1
(top left) A vehicle-mounted AMMOS unit and (top right) predetermined routes. Routes were mainly between the Lake Ontario shore and areas north of the highly urbanized parts of the greater Toronto area but could be extended west to the Milton area and north to Lake Simcoe as needed. The star marks the “home base” for the vehicles. (bottom) AMMOS data showing the rapid changes in temperature, dewpoint temperature, and relative humidity as the vehicle moved across the Lake Ontario lake-breeze front on the afternoon of 26 Jul 2015. Areas with blue shading are in lake air.
Citation: Bulletin of the American Meteorological Society 99, 5; 10.1175/BAMS-D-16-0162.1
(top left) A vehicle-mounted AMMOS unit and (top right) predetermined routes. Routes were mainly between the Lake Ontario shore and areas north of the highly urbanized parts of the greater Toronto area but could be extended west to the Milton area and north to Lake Simcoe as needed. The star marks the “home base” for the vehicles. (bottom) AMMOS data showing the rapid changes in temperature, dewpoint temperature, and relative humidity as the vehicle moved across the Lake Ontario lake-breeze front on the afternoon of 26 Jul 2015. Areas with blue shading are in lake air.
Citation: Bulletin of the American Meteorological Society 99, 5; 10.1175/BAMS-D-16-0162.1
Meteorological sensors were deployed on two buoys and two PA15 sailing committee boats. A standard WatchKeeper buoy monitored marine variables and used a cell modem to transmit observations every 10 min. A TRIAXYS directional wave buoy reported wave height, wave period, and wave direction every 30 min. For the boats, an AIRMAR magnetic flux compass was used to adjust the WXT520 winds for the changing orientation of the boats. Data were collected at 1-s intervals, and graphical products were generated and available on board, with data transmitted every minute via cell modems. Figure 4 shows an example of wind shifts of approximately 90-s duration and 15° of azimuth. These wind shifts have previously been qualitatively observed (Bethwaite 2010; D. Steenbergen 2015, personal communication) and are conjectured to be atmospheric waves that modulate the Ekman spiral wind profile and therefore result in slight wind direction changes (A. Hainsworth 2016, personal communication).
A 30-minute time series of 1-s wind (top) direction and (bottom) speed observed on a PA15 Sailing Committee boat (Port Credit Yacht Club, Heron II) showing a periodicity of about 90 s. Black dots are 1-s raw measurements, the red line is a 60-s running average, and the green line is a 600-s running average.
Citation: Bulletin of the American Meteorological Society 99, 5; 10.1175/BAMS-D-16-0162.1
A 30-minute time series of 1-s wind (top) direction and (bottom) speed observed on a PA15 Sailing Committee boat (Port Credit Yacht Club, Heron II) showing a periodicity of about 90 s. Black dots are 1-s raw measurements, the red line is a 60-s running average, and the green line is a 600-s running average.
Citation: Bulletin of the American Meteorological Society 99, 5; 10.1175/BAMS-D-16-0162.1
A 30-minute time series of 1-s wind (top) direction and (bottom) speed observed on a PA15 Sailing Committee boat (Port Credit Yacht Club, Heron II) showing a periodicity of about 90 s. Black dots are 1-s raw measurements, the red line is a 60-s running average, and the green line is a 600-s running average.
Citation: Bulletin of the American Meteorological Society 99, 5; 10.1175/BAMS-D-16-0162.1
In addition to the long-term AQ network covering the PA15 region, ECCC, MOECC, and UOT collaboratively operated four special sites to survey AQ. The focus was to investigate near-roadway pollutant levels to understand emissions and population exposures (NAPS 2017; Jeong et al. 2010; Wang et al. 2015). These sites were also the backbone for enhanced air pollution measurements and testing new methods for fine particulate matter <2.5 μm, physical and chemical characterization, and long-path turbulence and gaseous pollutant measurement over the dense traffic occurring on Highway 401 that runs through the middle of the city (You et al. 2017). Two of the sites, added during PA15 at Hanlan’s Point on the south side of Toronto Island and at ECCC headquarters in Downsview in northern Toronto, provided background measurements for the two urban stations at the UOT (downtown) and along Highway 401.
On board CRUISER, an ECCC mobile AQ laboratory, the latest technologies for detailed AQ research were deployed and provided unprecedented maps of pollutants such as nitrogen dioxide, benzene, hydrogen cyanide, ozone, and particulate black carbon (Levy et al. 2014). Mobile measurements, conducted on 13 days, revealed different spatial scales of the heterogeneity of urban air pollutants (Fig. 5), helped assess performance of the 2.5-km resolution AQ forecast model (GEM-MACHV2) and to assess the emissions inventory requirements needed for modeling. Experience from PA15 guided the planning for 23 additional sampling days conducted across the GTA and surrounding suburbs in September 2015 and January 2016. All these measurements were used to derive new maps of multipollutant exposures for use in future epidemiological research.
(top left) Composite of all CRUISER’s driving routes during and after PA15 to map multiple pollutant concentrations for GEM-MACH v2 evaluation. The grids in the inset map and the main figure represent the model grid squares (2.5 km). The main figure shows the PA15 benzene data (ppb) obtained using a proton transfer reaction time-of-flight mass spectrometer at 1-s time resolution, plotted along the west-to-east Highway 407 (green indicates low concentration; orange indicates high concentration). The measurements were segmented into 21 predefined 5-km sections for each passage along the highway (numbered in yellow boxes on the main figure). (bottom right) The percentiles derived from the 1-s data within each section were determined for each passage along the highway separately and then averaged to summarize the west-to-east variation in benzene concentrations. The 5th percentile (lower blue line) provides an indication of the urban background levels while the higher percentiles highlight parts of the city where benzene is systematically higher (hot spots). The regional background concentration, which was below the detection limit for benzene, was determined from the nearby semirural locations. The hot spot corresponds to an area with significant diesel truck and rail traffic. The west-to-east variation in the blue line (5th percentile) reveals how the urban background benzene builds up over the urbanized area but still exhibits spatial variability.
Citation: Bulletin of the American Meteorological Society 99, 5; 10.1175/BAMS-D-16-0162.1
(top left) Composite of all CRUISER’s driving routes during and after PA15 to map multiple pollutant concentrations for GEM-MACH v2 evaluation. The grids in the inset map and the main figure represent the model grid squares (2.5 km). The main figure shows the PA15 benzene data (ppb) obtained using a proton transfer reaction time-of-flight mass spectrometer at 1-s time resolution, plotted along the west-to-east Highway 407 (green indicates low concentration; orange indicates high concentration). The measurements were segmented into 21 predefined 5-km sections for each passage along the highway (numbered in yellow boxes on the main figure). (bottom right) The percentiles derived from the 1-s data within each section were determined for each passage along the highway separately and then averaged to summarize the west-to-east variation in benzene concentrations. The 5th percentile (lower blue line) provides an indication of the urban background levels while the higher percentiles highlight parts of the city where benzene is systematically higher (hot spots). The regional background concentration, which was below the detection limit for benzene, was determined from the nearby semirural locations. The hot spot corresponds to an area with significant diesel truck and rail traffic. The west-to-east variation in the blue line (5th percentile) reveals how the urban background benzene builds up over the urbanized area but still exhibits spatial variability.
Citation: Bulletin of the American Meteorological Society 99, 5; 10.1175/BAMS-D-16-0162.1
(top left) Composite of all CRUISER’s driving routes during and after PA15 to map multiple pollutant concentrations for GEM-MACH v2 evaluation. The grids in the inset map and the main figure represent the model grid squares (2.5 km). The main figure shows the PA15 benzene data (ppb) obtained using a proton transfer reaction time-of-flight mass spectrometer at 1-s time resolution, plotted along the west-to-east Highway 407 (green indicates low concentration; orange indicates high concentration). The measurements were segmented into 21 predefined 5-km sections for each passage along the highway (numbered in yellow boxes on the main figure). (bottom right) The percentiles derived from the 1-s data within each section were determined for each passage along the highway separately and then averaged to summarize the west-to-east variation in benzene concentrations. The 5th percentile (lower blue line) provides an indication of the urban background levels while the higher percentiles highlight parts of the city where benzene is systematically higher (hot spots). The regional background concentration, which was below the detection limit for benzene, was determined from the nearby semirural locations. The hot spot corresponds to an area with significant diesel truck and rail traffic. The west-to-east variation in the blue line (5th percentile) reveals how the urban background benzene builds up over the urbanized area but still exhibits spatial variability.
Citation: Bulletin of the American Meteorological Society 99, 5; 10.1175/BAMS-D-16-0162.1
UOT operated the MAPLE (Jeong et al. 2015) vehicular AQ laboratory. It focused on black carbon, nitrogen dioxide, and particles among other parameters and demonstrated the high variability along the major traffic highways within the urban environment (Jerrett et al. 2009).
Two supersites provided both surface and vertical profiling observations. The site at Pearson Airport was designed for aviation nowcasting (Isaac et al. 2014a), and the site at the University of Ontario Institute of Technology focused on the model parameterization of fog and cloud processes and satellite validation (Gultepe et al. 2014a,b, 2015, 2017).
Two Doppler lidars (Halo Photonics, Streamline Pro model) were deployed. The fixed lidar was positioned on the south shore of Toronto Island at the Hanlan’s Point AQ site. The second lidar was deployed on a truck and powered by a portable gas generator. Figure 6 illustrates the vertical velocities during an LB front passage on 28 July 2015.
Vertical velocity from (top) the Doppler lidars located at DOW (∼10 km inland and north of the shoreline) and (bottom) HAN during a lake-breeze passage. Gaps in the bottom panel indicate when the lidar was not vertically pointing. The surface wind data (not shown) indicated that the lake breeze occurred at around 2000 UTC at DOW, coinciding with the dominant vertical velocity pattern in the Doppler data. The lake breeze was not evident at HAN, which is at the south side of Toronto Island, indicating that the origin of the lake breeze was north of the station. See Figs. 1 and A2 for locations.
Citation: Bulletin of the American Meteorological Society 99, 5; 10.1175/BAMS-D-16-0162.1
Vertical velocity from (top) the Doppler lidars located at DOW (∼10 km inland and north of the shoreline) and (bottom) HAN during a lake-breeze passage. Gaps in the bottom panel indicate when the lidar was not vertically pointing. The surface wind data (not shown) indicated that the lake breeze occurred at around 2000 UTC at DOW, coinciding with the dominant vertical velocity pattern in the Doppler data. The lake breeze was not evident at HAN, which is at the south side of Toronto Island, indicating that the origin of the lake breeze was north of the station. See Figs. 1 and A2 for locations.
Citation: Bulletin of the American Meteorological Society 99, 5; 10.1175/BAMS-D-16-0162.1
Vertical velocity from (top) the Doppler lidars located at DOW (∼10 km inland and north of the shoreline) and (bottom) HAN during a lake-breeze passage. Gaps in the bottom panel indicate when the lidar was not vertically pointing. The surface wind data (not shown) indicated that the lake breeze occurred at around 2000 UTC at DOW, coinciding with the dominant vertical velocity pattern in the Doppler data. The lake breeze was not evident at HAN, which is at the south side of Toronto Island, indicating that the origin of the lake breeze was north of the station. See Figs. 1 and A2 for locations.
Citation: Bulletin of the American Meteorological Society 99, 5; 10.1175/BAMS-D-16-0162.1
A 14-station research lightning detection system (SOLMA) was deployed in the GTA to sample the three-dimensional structure of individual light strokes and flashes (Rison et al. 1999; Sills et al. 2014). The objectives of this system, through integration with dual-polarization radar data, were to better understand the microphysics related to lightning initiation, to develop nowcasting techniques for severe weather (heavy rain, strong winds, hail, and tornadoes) to improve lead time for warnings, and to evaluate new technologies used for operational lightning detection, including the GOES-based Geostationary Lightning Mapper (Goodman et al. 2013) and improved sensors for the CLDN (Burrows and Kochtubajda 2010). Figure 7 shows a comparison of data from SOLMA and CLDN.
Data from about 1.5-s duration from SOLMA. The CLDN reported a single cloud-to-ground lightning flash (yellow triangle symbol on all graphs) though SOLMA shows an intricate cloud-to-cloud pattern stretching out to over 30 km in length. (top) A time–height graph with data color coded by time for interpretation of (bottom left) the 2D ground map and (middle),(bottom right) the vertical projections.
Citation: Bulletin of the American Meteorological Society 99, 5; 10.1175/BAMS-D-16-0162.1
Data from about 1.5-s duration from SOLMA. The CLDN reported a single cloud-to-ground lightning flash (yellow triangle symbol on all graphs) though SOLMA shows an intricate cloud-to-cloud pattern stretching out to over 30 km in length. (top) A time–height graph with data color coded by time for interpretation of (bottom left) the 2D ground map and (middle),(bottom right) the vertical projections.
Citation: Bulletin of the American Meteorological Society 99, 5; 10.1175/BAMS-D-16-0162.1
Data from about 1.5-s duration from SOLMA. The CLDN reported a single cloud-to-ground lightning flash (yellow triangle symbol on all graphs) though SOLMA shows an intricate cloud-to-cloud pattern stretching out to over 30 km in length. (top) A time–height graph with data color coded by time for interpretation of (bottom left) the 2D ground map and (middle),(bottom right) the vertical projections.
Citation: Bulletin of the American Meteorological Society 99, 5; 10.1175/BAMS-D-16-0162.1
HIGH-RESOLUTION AND URBAN MODEL DEMONSTRATION.
A variety of models were run for PA15 in experimental mode (run on the operational computer on a medium-priority basis) and are summarized in Table 2. The common theme is high resolution and the physics and chemistry required to capture details of the weather, AQ, and health affecting the warning and alerting scales.
Summary of modeling systems. Note that 1) all models were run on the operational supercomputer on the following decreasing priority: operational mode, experimental quasi-operational mode, and experimental best effort (from user account), and 2) only operational models used in conjunction with PA15 models are listed.
Numerical weather forecasting
Integration of different predictive components is a crucial step toward weather forecasting improvement in highly urbanized areas (Grimmond et al. 2015). The GEM (Zadra et al. 2008; Girard et al. 2014) was operated in limited-area mode (GEM-LAM) at various resolutions and driven by RDPS (Fillion et al. 2010). The HRDPS is run four times a day on a pan-Canadian domain at 2.5-km resolution (Milbrandt et al. 2016) and includes advanced physics and soil moisture and surface temperature assimilation (Carrera et al. 2015).
Subkilometer (1 km and 250 m, GEM-LAM Urban) urban modeling systems were set up over the PA15 domain with inclusion of the TEB (Masson 2000). The configuration included an increase of the near-surface vertical resolution to capture small-scale atmospheric circulations, such as the LB interacting with topography- and urban-induced modification of the flow (Leroyer et al. 2014). Surface temperatures for the Great Lakes were prescribed using 2-km hourly output from a coupled lake–atmosphere forecasting system (Dupont et al. 2012) based on the NEMO model. Forecasts started at 0600 UTC were valid for 24 h. WBGT and UTCI heat stress indices were computed for the 1-km and 250-m models. These products were available at the forecast desk but also to the weather and health web portal, providing unprecedented high-resolution products for health authorities.
Evaluation of the urban modeling system was conducted with the PA15 mesonet. Objective evaluation (May–August 2015) showed improvements at the surface when compared with operational forecasts or with the 2.5-km HRDPS model and as verified for traditional meteorological variables of near-surface air temperature and dewpoint temperature, precipitation, and wind speed. This is attributed to the combined effects of high resolution and the inclusion of urban thermodynamic and aerodynamic effects. Precipitation statistics [>2 mm (6 h)−1] show an improvement in the probability of detection and in the reliability of precipitation forecasts with 250-m grid spacing (Fig. 8).
Objective evaluation of precipitation occurrence for the 250-m and 2.5-km GEM-LAM systems. (left) Probability of detection and (right) false alarm ratio for the category above 2 mm of precipitation per 6 h.
Citation: Bulletin of the American Meteorological Society 99, 5; 10.1175/BAMS-D-16-0162.1
Objective evaluation of precipitation occurrence for the 250-m and 2.5-km GEM-LAM systems. (left) Probability of detection and (right) false alarm ratio for the category above 2 mm of precipitation per 6 h.
Citation: Bulletin of the American Meteorological Society 99, 5; 10.1175/BAMS-D-16-0162.1
Objective evaluation of precipitation occurrence for the 250-m and 2.5-km GEM-LAM systems. (left) Probability of detection and (right) false alarm ratio for the category above 2 mm of precipitation per 6 h.
Citation: Bulletin of the American Meteorological Society 99, 5; 10.1175/BAMS-D-16-0162.1
Figure 9a shows an example from the 250-m urban GEM-LAM (0100 UTC 29 July 2015) where the warm air temperatures over the GTA were influenced by a persistent inland LB front, consistent with the mesoscale analysis (Fig. 9b). The magnitude and spatial variability of human thermal stress indices were evaluated using the mesonet (Leroyer et al. 2018).
(a) Evening near-surface temperature and wind vector maps of the 250-m urban GEM-LAM at 0100 UTC 29 Jul 2015. The black line is the forecasted lake-breeze front and the dashed white line is from the subjective mesoanalysis. (b) The meteorological station plots, radar data, and analyzed lake-breeze front [dashed white line in (a)].
Citation: Bulletin of the American Meteorological Society 99, 5; 10.1175/BAMS-D-16-0162.1
(a) Evening near-surface temperature and wind vector maps of the 250-m urban GEM-LAM at 0100 UTC 29 Jul 2015. The black line is the forecasted lake-breeze front and the dashed white line is from the subjective mesoanalysis. (b) The meteorological station plots, radar data, and analyzed lake-breeze front [dashed white line in (a)].
Citation: Bulletin of the American Meteorological Society 99, 5; 10.1175/BAMS-D-16-0162.1
(a) Evening near-surface temperature and wind vector maps of the 250-m urban GEM-LAM at 0100 UTC 29 Jul 2015. The black line is the forecasted lake-breeze front and the dashed white line is from the subjective mesoanalysis. (b) The meteorological station plots, radar data, and analyzed lake-breeze front [dashed white line in (a)].
Citation: Bulletin of the American Meteorological Society 99, 5; 10.1175/BAMS-D-16-0162.1
AQ modeling demonstration
ECCC generated high-resolution (2.5-km resolution) AQ forecasts (Stroud et al. 2011). This was the first test of ECCC’s next-generation AQ model (GEM-MACHV2) and included new mobile emission inputs for the GTA based on the recent Canadian on-road network database (Zhang et al. 2012a,b; Su et al. 2010). The advection and physics in the AQ model system were based on the HRDPS. The model ran successfully every day, generating 24-h products from the 0600 UTC model run by 0900 UTC. ECCC also generated a near-real-time high-resolution chemical objective analysis (also at 2.5-km grid spacing) for the first time [adapted after Robichaud and Ménard (2014)]. This product proved useful to forecasters in identifying model biases in near–real time. For example, modeled ozone concentrations were often overpredicted near the lake on sunny days with light winds in the afternoon.
In the late afternoon of 28 July 2015, the Toronto north AQ station and CRUISER mobile vehicle both measured 90 ppbv of ozone. Figure 10a shows the surface ozone predictions and winds at 2000 UTC, along with the vertical cross section from south to north (Fig. 10b). An LB front is evident from the model winds, and it runs parallel to the lakeshore, north of the urban core. The maximum predicted ozone (100–105 ppbv) is located between Vaughan and Brampton, which is in an area of surface wind convergence (see cross section) and the location of numerous transportation emission sources (including major highways, rail, and airport).
(a) GEM-MACHV2 predictions at 2000 UTC 28 Jul for surface ozone, (b) vertical cross section from south to north shown in (a) through points A, B, and C, and (c) the mesonet analysis for the lake-breeze front. Also included in (a) is a table with modeled and observed concentrations at three stations (points 1–3) and CRUISER (point 4). The dashed line is the diagnosed location of the lake-breeze front from (c). Orange lines are the locations of major highways. The Y axis in (b) is the altitude above ground using the model hybrid pressure–terrain coordinate system. Physically, the values of 1.0, 0.9, and 0.7 correspond to the surface, 1 km, and 3 km above the ground.
Citation: Bulletin of the American Meteorological Society 99, 5; 10.1175/BAMS-D-16-0162.1
(a) GEM-MACHV2 predictions at 2000 UTC 28 Jul for surface ozone, (b) vertical cross section from south to north shown in (a) through points A, B, and C, and (c) the mesonet analysis for the lake-breeze front. Also included in (a) is a table with modeled and observed concentrations at three stations (points 1–3) and CRUISER (point 4). The dashed line is the diagnosed location of the lake-breeze front from (c). Orange lines are the locations of major highways. The Y axis in (b) is the altitude above ground using the model hybrid pressure–terrain coordinate system. Physically, the values of 1.0, 0.9, and 0.7 correspond to the surface, 1 km, and 3 km above the ground.
Citation: Bulletin of the American Meteorological Society 99, 5; 10.1175/BAMS-D-16-0162.1
(a) GEM-MACHV2 predictions at 2000 UTC 28 Jul for surface ozone, (b) vertical cross section from south to north shown in (a) through points A, B, and C, and (c) the mesonet analysis for the lake-breeze front. Also included in (a) is a table with modeled and observed concentrations at three stations (points 1–3) and CRUISER (point 4). The dashed line is the diagnosed location of the lake-breeze front from (c). Orange lines are the locations of major highways. The Y axis in (b) is the altitude above ground using the model hybrid pressure–terrain coordinate system. Physically, the values of 1.0, 0.9, and 0.7 correspond to the surface, 1 km, and 3 km above the ground.
Citation: Bulletin of the American Meteorological Society 99, 5; 10.1175/BAMS-D-16-0162.1
The observed front [Figs. 10a (dashed line) and 10c] also runs parallel to the shoreline but does not penetrate as far inland, just reaching Brampton but south of Vaughan. The Brampton station (marked as point 3 in Fig. 10a) recorded only 51 ppbv, while the model predicted 80–85 ppbv. This difference occurred because the modeled front was north of the Brampton station while the observed front, the Toronto north station (marked as as point 1), the Burlington station (marked as point 2), and CRUISER (marked as point 4) were south. They measured 87, 84, and 73 ppbv while the model predicted 80–85, 75–80, and 65–70 ppbv, respectively. Thus, the model predicted reasonable concentrations for air masses on either side of the LB front (see table inset in Fig. 10a). However, the location of the LB front was predicted too far north. Figure 10b shows that the model predicted a classic LB circulation on this day.
Wave modeling
A wave forecast suite was developed using WAVEWATCH III, version 4.18 (Tolman and WAVEWATCH III Development Group 2014). The suite consisted of a control plus 20-member ensemble with grid spacing of 2.5-km covering the Great Lakes driven by forecast winds from the regional ensemble prediction system and two deterministic (1-km and 250-m grid spacing) wave prediction systems covering Lake Ontario. The ensemble was used to issue warning of the likelihood of wave events exceeding critical thresholds with lead times of up to 72 h [adapted after Bernier and Thompson (2015)]. The probabilities were calculated based on the number of members exceeding the predetermined threshold of 1 and 2 m. Deterministic forecasts were also used to refine the guidance with four daily updates at grid spacing of 1 km, and a daily update at 250 m driven by the GEM 250-m atmospheric forecast. At such resolution, the wind spatial variability can have considerable impacts on the wave forecast variability (Fig. 11).
The (a) 250-m and (b) 1-km significant wave height (colors) and wind speed (wind barbs). Wave heights from four buoys are shown in colored circles.
Citation: Bulletin of the American Meteorological Society 99, 5; 10.1175/BAMS-D-16-0162.1
The (a) 250-m and (b) 1-km significant wave height (colors) and wind speed (wind barbs). Wave heights from four buoys are shown in colored circles.
Citation: Bulletin of the American Meteorological Society 99, 5; 10.1175/BAMS-D-16-0162.1
The (a) 250-m and (b) 1-km significant wave height (colors) and wind speed (wind barbs). Wave heights from four buoys are shown in colored circles.
Citation: Bulletin of the American Meteorological Society 99, 5; 10.1175/BAMS-D-16-0162.1
Products were regularly updated at the forecast desk and on the ECPASS website. Preliminary validation against in situ buoys shows that both the ensemble and deterministic systems performed well and thus they have continued to produce experimental wave guidance after PA15.
NEXT-GENERATION FORECASTING AND NOWCASTING DEMONSTRATION.
A demonstration of prototype forecasting and nowcasting techniques using an object-based approach (Sills 2009) involved four ECCC meteorologists working at two research support desks (Sills and Taylor 2008). The meteorologists created MetObject1 analyses, nowcasts, and forecasts focused on operational severe and high-impact weather warnings and forecasts using iCAST software (Sills et al. 2009) and the Aurora workstation (Greaves et al. 2001). ECCC is investigating the use of this approach for operational warnings and high-impact warning–related forecasts and in particular of automated generation of high-quality “first guess” MetObjects that would reduce the need for forecaster modifications. This demonstration focused on the associated workload and workflows.
One desk focused on hourly analyses of features important for the development of thunderstorms and severe weather (mainly synoptic fronts and mesoscale boundaries such as LB fronts and thunderstorm gust fronts), while the other desk-generated probabilistic thunderstorm forecasts at 3-h intervals for day 1 (1800, 2000, and 0000 UTC) and 6-h intervals for days 2 and 3 (0600, 1200, 1800, and 0000 UTC). The MetObject analyses could be shared between the two desks (see Fig. 12) as necessary and collaboration was encouraged. They were available to the OSPC forecasters, PA15 forecasters, and briefers in real time.
Research support desk–generated MetObject-based products for 24 Jul 2015 including (top) a composite synoptic prognosis plot with forecaster-assessed fronts, troughs, jets, and mesoscale boundaries valid at 1800 UTC as well as thunderstorm and severe weather threat areas valid 1800–2100 UTC and (bottom) a composite mesoscale analysis plot showing forecaster-assessed lake-breeze fronts based on satellite imagery, radar data, and mesonet observations valid at 1800 UTC. Synoptic-scale features were identified using standard operational meteorology methods while mesoscale boundaries were identified using techniques described in Sills et al. (2011).
Citation: Bulletin of the American Meteorological Society 99, 5; 10.1175/BAMS-D-16-0162.1
Research support desk–generated MetObject-based products for 24 Jul 2015 including (top) a composite synoptic prognosis plot with forecaster-assessed fronts, troughs, jets, and mesoscale boundaries valid at 1800 UTC as well as thunderstorm and severe weather threat areas valid 1800–2100 UTC and (bottom) a composite mesoscale analysis plot showing forecaster-assessed lake-breeze fronts based on satellite imagery, radar data, and mesonet observations valid at 1800 UTC. Synoptic-scale features were identified using standard operational meteorology methods while mesoscale boundaries were identified using techniques described in Sills et al. (2011).
Citation: Bulletin of the American Meteorological Society 99, 5; 10.1175/BAMS-D-16-0162.1
Research support desk–generated MetObject-based products for 24 Jul 2015 including (top) a composite synoptic prognosis plot with forecaster-assessed fronts, troughs, jets, and mesoscale boundaries valid at 1800 UTC as well as thunderstorm and severe weather threat areas valid 1800–2100 UTC and (bottom) a composite mesoscale analysis plot showing forecaster-assessed lake-breeze fronts based on satellite imagery, radar data, and mesonet observations valid at 1800 UTC. Synoptic-scale features were identified using standard operational meteorology methods while mesoscale boundaries were identified using techniques described in Sills et al. (2011).
Citation: Bulletin of the American Meteorological Society 99, 5; 10.1175/BAMS-D-16-0162.1
The skill of human-generated forecasts relative to automated NWP-based guidance was evaluated. Real-time verification scores were available following each shift so that meteorologists could assess their performance while forecast decisions were still fresh. Verification statistics using data will be the focus of a future publication.
AUTOMATED VENUE NOWCASTING.
A blended nowcasting system for the 17 venue locations was demonstrated. Multiple NWP model forecasts (the 250-m, 1.0-km, 2.5-km, and 10-km model outputs) were bias corrected and combined using a dynamically adjusted weighting approach based on the observations by the INTW nowcast system (Huang et al. 2012). Updates were available every 10 min for real-time use with other forecasts and nowcasts (Fig. 13). The nowcasts from INTW were compared to the NWP models, to the operational scribe INCS (C. Landry et al. 2012, personal communication), and to persistence. Preliminary verification results (Huang 2016) show that for temperature, relative humidity, and wind speed, INTW performed considerably better than the other methods. INCS performed well for temperature but had very large errors for wind speed. Neither method performed well for wind direction.
INTW nowcasts for temperature, relative humidity, wind direction, and wind speed for the Royal Canadian Henley (RCH/Z4W) venue site on the north shore of the Niagara peninsula. The gray vertical line is T0 at 1640 UTC. To the left of T0, INTW 1-h nowcasts are compared against observations. To the right of T0, INTW nowcasts generated in 10-min time steps out to 8 h and valid for 1640 UTC are shown. Forecasts are from the RDPS (10 km) and HRDPS (2.5 km, 1.0 km, and 250 m). INCS nowcasts were not available in real time but are shown for comparison. In this particular case, all models were predicting the Lake Ontario lake breeze would affect the site near 1400 UTC (wind shift to north-northwest), but observations show that the lake breeze did not arrive there before 1640 UTC (wind remaining from the south). Later observations indicate that the lake breeze did, however, arrive at the station by 1800 UTC (see Fig. 12).
Citation: Bulletin of the American Meteorological Society 99, 5; 10.1175/BAMS-D-16-0162.1
INTW nowcasts for temperature, relative humidity, wind direction, and wind speed for the Royal Canadian Henley (RCH/Z4W) venue site on the north shore of the Niagara peninsula. The gray vertical line is T0 at 1640 UTC. To the left of T0, INTW 1-h nowcasts are compared against observations. To the right of T0, INTW nowcasts generated in 10-min time steps out to 8 h and valid for 1640 UTC are shown. Forecasts are from the RDPS (10 km) and HRDPS (2.5 km, 1.0 km, and 250 m). INCS nowcasts were not available in real time but are shown for comparison. In this particular case, all models were predicting the Lake Ontario lake breeze would affect the site near 1400 UTC (wind shift to north-northwest), but observations show that the lake breeze did not arrive there before 1640 UTC (wind remaining from the south). Later observations indicate that the lake breeze did, however, arrive at the station by 1800 UTC (see Fig. 12).
Citation: Bulletin of the American Meteorological Society 99, 5; 10.1175/BAMS-D-16-0162.1
INTW nowcasts for temperature, relative humidity, wind direction, and wind speed for the Royal Canadian Henley (RCH/Z4W) venue site on the north shore of the Niagara peninsula. The gray vertical line is T0 at 1640 UTC. To the left of T0, INTW 1-h nowcasts are compared against observations. To the right of T0, INTW nowcasts generated in 10-min time steps out to 8 h and valid for 1640 UTC are shown. Forecasts are from the RDPS (10 km) and HRDPS (2.5 km, 1.0 km, and 250 m). INCS nowcasts were not available in real time but are shown for comparison. In this particular case, all models were predicting the Lake Ontario lake breeze would affect the site near 1400 UTC (wind shift to north-northwest), but observations show that the lake breeze did not arrive there before 1640 UTC (wind remaining from the south). Later observations indicate that the lake breeze did, however, arrive at the station by 1800 UTC (see Fig. 12).
Citation: Bulletin of the American Meteorological Society 99, 5; 10.1175/BAMS-D-16-0162.1
HEALTH INITIATIVE.
PA15 represented a unique opportunity to develop a portfolio of activities designed to enhance existing weather- and health-related service offerings (such as the air quality health index and heat warning and information system) and to utilize partnerships, technology, and expertise to forge possible new service directions. Over 30 different groups representing federal, provincial, and municipal governments, local public health, academia, public education, and the private sector cooperated in developing numerous initiatives and their demonstration, leveraging the PA15 observations, modeling, and forecasting assets. The following is a brief description of only a few such activities [see Johnston et al. (2017), provided as supplemental information, for details].
PA15 served as a catalyst for the implementation of a new health evidence-based heat warning information system. This impact-based service was developed over 2.5 years, in collaboration with HC, PHO, MHLTC, and local PHUs. An integral part of this multiagency undertaking was to shift the ECCC heat warning criteria from those based on climatology to those based on an epidemiological analysis from PHO. Using these new multiregional warning criteria, SOPs were developed to create an early notification service for public health decision-makers and community partners. These services supported early mobilization of community partners to reduce heat–health risk, especially for those most vulnerable. Under the collaboration, the MHLTC established similar SOPs to support the PHU to use the new early notification service and warnings and to harmonize communications and response actions. New streamlined health messages, developed by HC and vetted by the partners (above), provided impacts and call-to-action statements to reduce heat–health risk. Those messages were integrated into the public heat warnings issued by the forecasters at the OSPC, 18–24 h in advance of the heat event. This integrated alert/response approach tested during PA15 subsequently served as the model for the modernization of heat warnings in other Canadian provinces.
The AQHI was implemented in Ontario for the summer of 2015 in collaboration with MOECC and MHLTC (Stieb et al. 2008; ECCC 2017a). A feature of the Ontario implementation was the development of public alerts in the form of special AQ statements or smog and air health advisories when either the AQHI was forecast for high risk or forecast concentrations of ground-level ozone exceeded provincial guidelines. The two additional stations contributed to a roadside monitoring initiative (described above) and increased the number of forecast locations to eight. It also allowed for public testing of a new rolling 1-h AQHI forecast (for the next 18 h).
New low-cost, real-time UV instrumentation was deployed and tested along with UV alerts when the UV index was forecast to reach 8 or above. A new UV index model (Tereszchuk et al. 2017) was also implemented using assimilation of satellite column ozone data and forecasts of UV spectral band irradiances (Li and Barker 2005) from a development version of GEM with linearized ozone chemistry (de Grandpré et al. 2016). The current ECCC operational UV index model (Burrows et al. 1994; He et al. 2013) relies on a statistical and empirical approach using weather conditions and sparsely distributed ground-based column ozone measurements. An assessment of the feasibility and timeline for augmenting the existing service with the enhanced model is currently underway.
A weather and health decision-support portal for public health units in the PA15 area was established in collaboration with KFLA. The ECCC Alert-Me service, a push email service, was enhanced with health-related alerts and links to the portal. All forecasts, alerts, mesonet data, and enhanced modeling data for risk factors (heat, AQ, and UV) were made available geospatially, in real time, to support decision-making (Fig. 14). The functionality built for PA15 and much of the ECCC data were transferred to the KFLA’s Public Health Information System at the conclusion of PA15, as will be the threshold-based alerting for AQHI through ECCC’s Alert-Me service.
Composite image from the weather and health decision-support portal showing overlays of real-time radar, alert layer, and mesonet data.
Citation: Bulletin of the American Meteorological Society 99, 5; 10.1175/BAMS-D-16-0162.1
Composite image from the weather and health decision-support portal showing overlays of real-time radar, alert layer, and mesonet data.
Citation: Bulletin of the American Meteorological Society 99, 5; 10.1175/BAMS-D-16-0162.1
Composite image from the weather and health decision-support portal showing overlays of real-time radar, alert layer, and mesonet data.
Citation: Bulletin of the American Meteorological Society 99, 5; 10.1175/BAMS-D-16-0162.1
University of California, San Diego, and Texas Tech University researchers obtained EMS ambulance data to explore in detail the effects of heat on human health. Normally these types of studies are restricted to only a single observation station for the entire urban area. Research will focus on determining which heat stress index/model—the WBGT, Humidex, or the physically based COMFA model (Brown and Gillespie 1986; Vanos et al. 2012; Graham et al. 2016)—is able to best predict heat-related EMS calls. Figure 15 shows the variability of the average of the daily maximum WBGT (Herdt 2017), which allows for preparing strategic public health response plans (e.g., cooling station placement) during extreme heat events. The maximum instantaneous value of 27.8°C was measured during the men’s and women’s soccer matches in Hamilton on 26 July and coincided with a maximum in heat-related emergency response calls.
Station-specific WBGT index values (average of the daily maximum hourly value; °C) across the games area. Each circle represents an individual station; color and size reflect the magnitude of the WBGT, with darker and larger circles indicating a higher value. This shows the variability of the WBGT. Normally only a single value is available.
Citation: Bulletin of the American Meteorological Society 99, 5; 10.1175/BAMS-D-16-0162.1
Station-specific WBGT index values (average of the daily maximum hourly value; °C) across the games area. Each circle represents an individual station; color and size reflect the magnitude of the WBGT, with darker and larger circles indicating a higher value. This shows the variability of the WBGT. Normally only a single value is available.
Citation: Bulletin of the American Meteorological Society 99, 5; 10.1175/BAMS-D-16-0162.1
Station-specific WBGT index values (average of the daily maximum hourly value; °C) across the games area. Each circle represents an individual station; color and size reflect the magnitude of the WBGT, with darker and larger circles indicating a higher value. This shows the variability of the WBGT. Normally only a single value is available.
Citation: Bulletin of the American Meteorological Society 99, 5; 10.1175/BAMS-D-16-0162.1
SUMMARY.
A high-level overview of the ECPASS project has been presented. This was initiated as an internal project to showcase the research and innovations within ECCC. The prestige of PA15; the severe weather, AQ, and health warning challenges at urban scales; and the firm deadlines provided a natural motivation to coordinate, to focus normally separate research activities, and to accelerate the technology transfer process. There is much specific research to be done and include the following: urban meteorological monitoring and its data quality control; high-resolution assimilation and modeling; understanding LB characteristics and their impact on severe weather; high-resolution AQ analysis for model initialization; risk communication of high-resolution weather, air quality, and heat warnings; societal impacts studies; development of wind-wave stress parameterizations using wind profiles derived from the Doppler lidar and wave buoy observations; investigation of the appropriate turbulence scheme for different model resolutions; characterization of the urban environment on weather; and the effective monitoring and impact of roads and heavy car traffic on air quality.
It is abundantly clear that there must be a match in scale (spatial and temporal) among the observations, the model, and the model physics; otherwise, the predictions are unverifiable. Developments in monitoring technologies, data sharing, observation standards, and appropriate data quality analysis at high resolution are needed. Model improvements include better representation of the underlying urban surface, increased vertical resolution in the boundary layer, and better understanding and representation of physical processes and high-resolution assimilation techniques. There must also be a match in scale between the services and the predictions; otherwise, it is not an efficient use of resources. Many of the end users are far from being able to use the high-resolution predictions. Engagement with end users, perhaps in a test bed environment, is needed to bridge this gap.
The mesonet initiative acquired data, for the first time, from other agencies through the operational data management system. Observational legacies include three additional surface stations, the capability to operationally collect 1-min data, and new definitions for wind gust, leading to a better understanding of meteorological and chemical processes. The mesonet data were supplemented with black globe temperature sensors, which were complemented by high-resolution heat stress model predictions (1-km resolution out to 24 h) that represented a revolution for health authorities.
The compact stations performed well in the summer environment. Small, low-cost AQ sensors were deployed and tested and now routinely provide new information on the variability of various chemical species in the urban environment. Doppler lidars provided low-level wind and vertical motion information. A new AQ monitoring site was permanently established at ECCC headquarters. CRUISER data from 13 driving days during PA15 and 23 additional days (post PA15) lead to improved GEM-MACHV2 parameterizations and also to new empirical models of urban-scale air pollutant exposures for use in epidemiological research (Stroud et al. 2014; Moussa et al. 2016).
The data from PA15 validated the environmental and health prediction systems and enable their rapid technology transfer from research to operations. GEM-MACHV2 is now the operational version of Canada’s operational AQ deterministic forecasting system (Makar et al. 2015).
It also supported the development of the first-ever Great Lakes ensemble wave forecast system. In addition, high-resolution deterministic systems were used to study the impact of wind spatial variability on the wave fields. The Great Lakes deterministic and ensemble forecast systems are expected to become operational in the near future.
The forecast process using MetObjects was demonstrated to be effective and led to service production efficiencies. Nowcasting using the combination of observations and multiscale models using the INTW system showed that it outperformed current operational techniques.
PA15 was invaluable in galvanizing partner/stakeholder interest and creating the internal momentum to strengthen existing and introduce new weather-related health services. It served as a catalyst for implementing the AQHI in the province and associated reporting/forecasting and launched a new health evidence-based heat warning service. In addition, demonstration/showcase activities included the use of e-mail technology to push alerts to decision-makers and testing of a new UV index, which will underpin future national service offerings. The technology and expertise tested during the PA15 has not only improved ECCC’s relationship with the health community but served to sharpen the justification for modernizing decision support tools for temperature warnings. From a health services perspective, the experience of PA15 has had a direct, positive, and lasting legacy.
The data from the project has been preserved for sharing through the Government of Canada open data portal (ECCC 2017b). There is much to be explored and investigated. The focus of the project was on the LB. However, quantifying the urban impact on the development of the LB is just beginning to be explored in the 250-m GEM-LAM through the mesonet and via Doppler lidar data (Mariani et al. 2018). Analysis with proprietary emergency services data allows for a detailed exploration of the effects of heat on human health. Such capability is a revolution and will take time for societal benefits to be realized.
It is estimated that 70% of the world’s population will reside in urban environments by 2050, and the greatest change will be in developing and least developed countries (Zhu et al. 2012). Health will be dependent on the health system and urban infrastructure (www.canue.ca) but also on the environment (Villeneuve et al. 2013; Smith et al. 2014; Crouse et al. 2015). Future projects will require even greater integration of meteorology, AQ, and health sciences observational and modeling tools.
In PA15, the responsibility for AQ and health warnings and research are at multiple levels of government and with different agencies within each level. The focus on urban scales is implicitly high resolution (hundreds of meters), which is emerging in operational atmospheric numerical weather prediction and within the expertise of national meteorological and hydrological services. It is also required to adequately capture the impact of major roadways and industries on air quality and on the impact of urban flooding. The integration of these services is emerging in cities throughout the globe (WMO 2017). Given that the efficient and effective provision of urban services are interrelated and often the jurisdiction of different agencies, leadership, cooperation, and collaboration among different levels of government and different scientific sectors is therefore necessary. The challenges are diverse and local to each city, and as the concepts are nascent, a comparative study is recommended in order to identify underlying core concepts that can be shared globally for efficient implementation.
THE DATA LEGACY
The operational data (mesonet, upper air, and air quality) from the project are available on the Government of Canada’s data portal (ECCC 2017b) and the NAPS data portal. Research datasets will appear as they are quality controlled. Quicklook products (images, movies) from radar and lidar are available. An hourly analysis of the positions of LB fronts (with accuracy of about 1 km; Sills et al. 2011) and other weather features was done. For selected periods of interest, the analysis was done at time intervals down to 5 min.
The availability of the legacy dataset was announced at a full-day event organized in conjunction with the International Workshop on Air Quality Forecasting (Stroud et al. 2017) in an effort to highlight the synergistic efforts of scientists from different disciplines required to understand and forecast the urban environment, share the lesson learned, and inspire similar initiatives worldwide. The event was capped with a panel discussion on the directions of urban environmental prediction systems, current challenges, and opportunities to address them. The panel underlined the need for coordinated studies in different urban environments to maximize the scientific relevance in light of the costs of working in such settings and highlighted gaps in such aspects as data assimilation and verification at the urban scale, all of which require good coverage from multiple sensors and observational platforms. Panel members called for the need for a rapid evolution of the scientific capacity to adapt and exploit technological advances from low-cost sensors to probabilistic modeling, a need driven by a growing appetite for information at a scale and altitude where it directly matters to citizens.
ACKNOWLEDGMENTS
Setting up a project involves the passion, competence, and hard work of many people. Administration, procurement, selecting mesonet sites, negotiating contracts, modifying operational procedures and processes, technology transfer to operations, and communicating the use of the new products are all significant challenges. It is estimated that over 350 people were involved in the PA15 project. The original proposal team consisted of operational and research staff and included Pat King, Sarah Wong, Ron Lee, and the lead author. Tsoi Yip was the driver behind the ECPASS concept. The informatics support from Delroy Barrett (operational data processing and management) and Dennis Wintjes (PA15 informatics infrastructure) was extraordinary. We would also like to recognize the cooperation of the MOECC, TRCA, GRCA, UOIT, Toronto Pubic Health, Ryerson University, the Toronto Port Authority, the PA15 Sailing Organizers, and the National Weather Service. The collaboration with Dave Steenbergen, vice-commodore for the PA15 sailing competition, was key in deploying the boat sensors. Thanks to Steve Goodman for arranging rapid-scan GOES data. Wiechers and Voogt were funded by an NSERC Discovery Grant (194105) awarded to J. A. Voogt.
APPENDIX: ACRONYMS.
| AMMOS | Automated Mobile Meteorological Observation System |
| AQ | Air quality |
| AQHI | Air quality health index |
| ATMOS | Atmospheric Transportable Meteorological Observing System |
| AURORA | Name of a workstation and application software |
| BGT | Black globe thermometer |
| BRF | Broadband radiative flux |
| CaLDAS | Canadian Land Data Assimilation System |
| CLDN | Canadian Lightning Detection Network |
| COMFA | Comfort Formula (a human heat balance model developed in southern Ontario and tested in Toronto) |
| CRUISER | Canadian Regional Urban Investigation System for the Environment and Research |
| DAS | Data acquisition system |
| DMA | Differential mobility analyzer |
| DND | Department of National Defense |
| DOW | Downsview ECCC headquarters |
| ECCC | Environment and Climate Change Canada (formerly Environment Canada) |
| ECPASS | ECCC Pan and Parapan American Science Showcase |
| EF | Enhanced Fujita (tornado rating) |
| EMS | Emergency Management Services |
| EnVar | Ensemble–variational |
| FAR | False alarm rate |
| FMD | Fog measuring device |
| FSOS | Fog and Snow Observing System |
| FTIR | Fourier transform infrared spectrometer |
| GAW | Global Atmospheric Watch |
| GCIP | Ground Cloud Imaging Probe |
| GDPS | Global Deterministic Prediction System |
| GEM | Global Environmental Model |
| GEM-LAM | GEM Local Area Mode |
| GEM-MACHV1.5.1 | GEM Model Atmospheric Chemistry Version 1.5.1 |
| GEM-MACHV2 | GEM Model Atmospheric Chemistry Version 2 |
| GEM-CGL | GEM Coupled Great Lakes with NEMO |
| GLEWPS | Great Lakes Ensemble Wave Prediction System |
| GLDWPS | Great Lakes Deterministic Wave Prediction System |
| Pa15DWPS | PA15 Deterministic Wave Prediction Systems |
| GRCA | Grand River conservation area |
| GTA | Greater Toronto area |
| GOC | Government of Canada |
| GOES | Geostationary Operational Environmental Satellite |
| GPS | Global positioning system |
| GURME | GAW Urban Research Meteorology and Environment |
| HAN | Hanlan’s Point |
| HC | Health Canada |
| HRDPS | High Resolution Deterministic Prediction System |
| iCAST | Interactive Convective Analysis and Storm Tracking |
| INCS | Integrated Nowcasting System |
| INTW | Integrated nowcasting by weighting |
| IOGAPS | Integrated organic gas and particle sampler |
| IWAQFR | International Workshop on Air Quality Forecasting and Research (science workshop) |
| KFL&A | Kingston Frontenac Lexington and Addision (local health authority) |
| LB | Lake breeze |
| LWC | Liquid water content |
| LMA | Lightning Mapping Array |
| MAP | MetOne aerosol profiler |
| MAPLE | Mobile Analysis of Particulate in the Environment |
| MHLTC | Ontario Ministry of Health and Long-Term Care |
| MOB | Mobile lidar |
| MOECC | Ontario Ministry of Environment and Climate Change |
| MOH | Ministry of Health (provincial) |
| NAPS | National Air Pollution Surveillance |
| NAVCAN | Navigation Canada |
| NEMO | Nucleus for European Modelling of the Ocean |
| NOAA | National Oceanic and Atmospheric Administration |
| NoN | Network of Networks |
| NWP | Numerical weather prediction |
| OR | Outgoing radiation |
| OSPC | Ontario Storm Prediction Centre |
| P | Pressure |
| PA | Precipitation accumulation |
| PA15 | Pan and Parapan Games 2015 |
| PAH | Polycyclic aromatic hydrocarbons |
| PARSIVEL | Particle Size Velocity |
| PAX | Photoacoustic extinctometer |
| PHO | Public Health Ontario |
| PHU | Public health unit |
| PIR | Precision infrared radiometer |
| PM | Particulate matter |
| PM2.5 | Particulate matter less than 2.5 µm |
| POD | Probability of detection |
| PPI | Plan position indicator |
| PTR-TOF-MS | Proton transfer reaction time-of-flight mass spectrometer |
| PUMS | PanAm UOIT Meteorological Supersite |
| qv | Water vapor mixing ratio |
| RDPS | Regional Deterministic Prediction System |
| RH | Relative humidity |
| RHI | Range–height indicator |
| SAQS | Special air quality statements |
| SHARP | Synchronized Hybrid Ambient, Real-Time Particulate |
| SI-131 | Sensor infrared model 131 |
| SOLMA | Southern Ontario Lightning Mapping Array |
| SOP | Standard operating procedure |
| SP2 | Single-particle soot photometer |
| SP-TOF | Soot particle-time of flight |
| ST | Sensor temperature |
| T | Temperature |
| TEB | Town Energy Balance Scheme |
| TRCA | Toronto and Region Conservation Authority |
| UOIT | University of Ontario Institute of Technology |
| UOT | University of Toronto |
| UTCI | Universal thermal and climate index |
| UV | Ultraviolet |
| VPR | Vertically pointing radar |
| VOC | Volatile organic compound |
| WBGT | Wet-bulb globe temperature, computed from model |
| WMO | World Meteorological Organization |
| WMS | Web mapping service |
| WSC | Water Survey of Canada |
| WU | Western University |
| WW3 | Wave Watch 3 lake circulation model |
APPENDIX: METEOROLOGICAL OVERVIEW.
During 10 July–15 August, southern Ontario experienced fairly typical summer weather. Maximum temperatures exceeded 30°C on about one-third of the days (Table A1), and heat warnings (Table A2) were issued for 18–19 and 27–30 July. Nearshore lake temperatures in the PA15 area ranged mainly between 17° and 24°C, though there was a period of upwelling alongshore near Toronto from 30 July to 9 August that resulted in water temperatures there as low as 12°C.
Observed values from Toronto Pearson Airport for both Jul and Aug compared to 1981–2010 Canadian climate normals.
The occurrence statistics for various conditions/events during the games period. Days are defined in local time, and the area considered is within a 100-km radius of Toronto Pearson Airport.
AQHI reached “moderate” on about one-third of the days and reached high only twice on 12 July (broad scale) and again in Toronto and Burlington on 28 July (urban-scale event) when special air quality statements were issued.
Thunderstorms developed on roughly half of the days during the PA15 period. Severe thunderstorm watches and warnings were issued on 19 July, 2 August, and 14 August. On 2 August, storms with widespread damaging winds, large hail, and tornadoes (four, ranging from EF1 to EF2) were also verified. A rainfall warning was issued for the northern part of the PA15 area on 14 July.
LBs from Lake Ontario, Lake Erie, Lake Simcoe, Georgian Bay, and even Lake Huron occurred on an almost daily basis in the PA15 area (34 of 37 days). Multiple days of LB activity are common in the Great Lakes region (Sills et al. 2011). There was an impressive 21-day episode from 18 July to 11 August.
APPENDIX: OBSERVATION SYSTEM INSTRUMENTATION.
This appendix provides a summary of the various instrument platforms. Figure A1 shows the surface monitoring observations officially available to operational forecasters before (left), during (middle), and after (right) PA15. The after-PA15 network (legacy) consisted of three new stations and existing stations from PA15 partners but now included into the operational monitoring system (see Table 1). In addition, the data can be collected every minute. See Fig. A2 for the location of the roadside monitoring and supersites (Tables A3–A13), which continue to operate as ongoing research sites.
Surface station locations before, during, and after PA15, showing the legacy for the operational forecasting system. The data management system is also capable of processing 1-min data. Black dots are existing meteorological sites. Magenta triangles are existing NAPS sites. Black diamonds are operational radar sites. Magenta squares are roadside initiative air quality sites. Red dots, squares, and crosses are research sites (ATMOS, LMA, and supersites, respectively). Cyan dots are compact sites. Yellow triangles are UV sites. Black and red stars are operational and research buoys, respectively.
Citation: Bulletin of the American Meteorological Society 99, 5; 10.1175/BAMS-D-16-0162.1
Surface station locations before, during, and after PA15, showing the legacy for the operational forecasting system. The data management system is also capable of processing 1-min data. Black dots are existing meteorological sites. Magenta triangles are existing NAPS sites. Black diamonds are operational radar sites. Magenta squares are roadside initiative air quality sites. Red dots, squares, and crosses are research sites (ATMOS, LMA, and supersites, respectively). Cyan dots are compact sites. Yellow triangles are UV sites. Black and red stars are operational and research buoys, respectively.
Citation: Bulletin of the American Meteorological Society 99, 5; 10.1175/BAMS-D-16-0162.1
Surface station locations before, during, and after PA15, showing the legacy for the operational forecasting system. The data management system is also capable of processing 1-min data. Black dots are existing meteorological sites. Magenta triangles are existing NAPS sites. Black diamonds are operational radar sites. Magenta squares are roadside initiative air quality sites. Red dots, squares, and crosses are research sites (ATMOS, LMA, and supersites, respectively). Cyan dots are compact sites. Yellow triangles are UV sites. Black and red stars are operational and research buoys, respectively.
Citation: Bulletin of the American Meteorological Society 99, 5; 10.1175/BAMS-D-16-0162.1
Location of roadside initiative (UOT and 401), roadside initiative background stations (DOW and HAN), Pearson Airport supersite (YYZ), and University of Ontario Institute of Technology supersite (UOIT) against a simplified land-use background. The Doppler lidars were located at HAN and DOW (when not mobile). Gray and brown are urban settlements, shades of green and yellow are rural, and blue is water.
Citation: Bulletin of the American Meteorological Society 99, 5; 10.1175/BAMS-D-16-0162.1
Location of roadside initiative (UOT and 401), roadside initiative background stations (DOW and HAN), Pearson Airport supersite (YYZ), and University of Ontario Institute of Technology supersite (UOIT) against a simplified land-use background. The Doppler lidars were located at HAN and DOW (when not mobile). Gray and brown are urban settlements, shades of green and yellow are rural, and blue is water.
Citation: Bulletin of the American Meteorological Society 99, 5; 10.1175/BAMS-D-16-0162.1
Location of roadside initiative (UOT and 401), roadside initiative background stations (DOW and HAN), Pearson Airport supersite (YYZ), and University of Ontario Institute of Technology supersite (UOIT) against a simplified land-use background. The Doppler lidars were located at HAN and DOW (when not mobile). Gray and brown are urban settlements, shades of green and yellow are rural, and blue is water.
Citation: Bulletin of the American Meteorological Society 99, 5; 10.1175/BAMS-D-16-0162.1
University of Western Ontario mobile weather station. Note that the viewing angle for the road-facing infrared radiometer was 40° (relative to nadir view = 0°), and radiative surface canyon wall temperatures may include a mix of sky and other obstructions while the station is in motion. All air temperature sensors were placed inside aspirated radiation shields.
Roadside monitoring at MOECC’s station near Highway 401.
ECCC long baseline roadside monitoring at Highway 401 site.
Roadside monitoring at UOT in downtown Toronto.
Roadside monitoring background by UOT on Toronto Island (HAN) site.
Roadside monitoring background by ECCC at DOW site.
ECCC CRUISER mobile air pollutant measurements.
University of Toronto MAPLE Measurements. MAPLE was deployed on 6 days during 21 Jul–5 Aug 2015.
ECCC Pearson Airport supersite (YYZ).
ECCC PUMS supersite at UOIT.
Satellite sites (FSOS towers) around UOIT supersite (∼1 km apart from PUMS). See Table A12 for sensor details.
REFERENCES
Alexander, L. S., 2012: Mesoscale boundaries and storm development in southwestern Ontario during ELBOW 2001. Ph.D. dissertation, York University, 325 pp. [Available from the Centre for Research in Earth and Space Science, York University, 4700 Keele St., Toronto, ON M3J 1P3, Canada.]
Bernier, N. B., and K. R. Thompson, 2015: Deterministic and ensemble storm surge prediction for Atlantic Canada with lead times of hours to ten days. Ocean Modell., 86, 114–127, https://doi.org/10.1016/j.ocemod.2014.12.002.
Bethwaite, F., 2010: High Performance Sailing: Faster Racing Techniques. 2nd ed. Bloomsbury, 448 pp.
Boodoo, S., D. Hudak, A. Ryzhkov, P. Zhang, N. Donaldson, D. Sills, and J. Reid, 2015: Quantitative precipitation estimation from a C-band dual-polarized radar for the 8 July 2013 flood in Toronto, Canada. J. Hydrometeor., 16, 2027–2044, https://doi.org/10.1175/JHM-D-15-0003.1.
Brook, J. R., P. A. Makar, D. M. L. Sills, K. L. Hayden, and R. McLaren, 2011: Exploring the nature of air quality over southwestern Ontario: Main findings from the Border Air Quality and Meteorology Study. Atmos. Chem. Phys., 13, 10 461–10 482, https://doi.org/10.5194/acp-13-10461-2013.
Brown, R. D., and T. J. Gillespie, 1986: Estimating outdoor thermal comfort using a cylindrical radiation thermometer and an energy budget model. Int. J. Biometeor., 30, 43–52, https://doi.org/10.1007/BF02192058.
Burrows, W. R., and B. Kochtubajda, 2010: A decade of cloud-to-ground lightning in Canada: 1999–2008. Part 1: Flash density and occurrence. Atmos.–Ocean, 48, 177–194, https://doi.org/10.3137/AO1118.2010.
Burrows, W. R., M. Vallée, D. I. Wardle, J. B. Kerr, L. J. Wilson, and D. W. Tarasick, 1994: The Canadian operational procedure for forecasting total ozone and UV radiation. Meteor. Appl., 1, 247–265, https://doi.org/10.1002/met.5060010307.
Carrera, M. L., S. Belair, and B. Bilodeau, 2015: The Canadian Land Data Assimilation System (CaLDAS): Descripton and synthetic evaluation study. J. Hydrometeor., 16, 1293–1314, https://doi.org/10.1175/JHM-D-14-0089.1.
Crouse, D. L., and Coauthors, 2015: Within- and between-city contrasts in nitrogen dioxide and mortality in 10 Canadian cities; a subset of the Canadian Census Health and Environment Cohort (CanCHEC). J. Exposure Sci. Environ. Epidemiol., 25, 482–489, https://doi.org/10.1038/jes.2014.89.
de Grandpré, J., M. Tanguay, A. Qaddouri, M. Zerroukat, and C. A. McLinden, 2016: Semi-Lagrangian advection of stratospheric ozone on a Yin–Yang grid system. Mon. Wea. Rev., 144, 1035–1050, https://doi.org/10.1175/MWR-D-15-0142.1.
Doyle, C., 2010: Weather forecasts for the 2010 Winter Olympic Games—Pre-games challenges and user requirements in an El Niño season. 14th Conf. on Mountain Meteorology, Olympic Valley, CA, Amer. Meteor. Soc., 15.1, https://ams.confex.com/ams/14MountMet/webprogram/Paper170046.html.
Duan, Y., and Coauthors, 2012: An overview of the Beijing 2008 Olympics Research and Development Project (B08RDP). Bull. Amer. Meteor. Soc., 93, 381–403, https://doi.org/10.1175/BAMS-D-11-00115.1.
Dupont, F., P. Chittibabu, V. Fortin, Y. R. Rao, and Y. Lu, 2012: Assessment of NEMO-based hydrodynamic modelling system for the Great Lakes. Water Quality Res. J., 47, 198–214, https://doi.org/10.2166/wqrjc.2012.014.
ECCC, 2016: Toronto 2015 Pan and Parapan American Games—An Environment and Climate Change Canada Perspective. WMO Bull., 65, 42–47.
ECCC, 2017a: Air quality health index. Accessed 15 Mar 2018, www.canada.ca/en/environment-climate-change/services/air-quality-health-index.html.
ECCC, 2017b: TO2015 Pan and Parapan American Games atmospheric monitoring observatios. Accessed 15 Mar 2018, http://donnees.ec.gc.ca/data/weather/monitor/to2015-pan-and-parapan-american-games-atmospheric-monitoring-observations/.
Fillion, L., and Coauthors, 2010: The Canadian Regional Data Assimilation and Forecasting System. Wea. Forecasting, 25, 1645–1669, https://doi.org/10.1175/2010WAF2222401.1.
Girard, C., and Coauthors, 2014: Staggered vertical discretization of the Canadian Environmental Multiscale (GEM) model using a coordinate of the log-hydrostatic-pressure type. Mon. Wea. Rev., 142, 1183–1196, https://doi.org/10.1175/MWR-D-13-00255.1.
Goodman, S. J., and Coauthors, 2013: The GOES-R Geostationary Lightning Mapper (GLM). Atmos. Res., 125–126, 34–49, https://doi.org/10.1016/j.atmosres.2013.01.006.
Graham, A., R. D. Brown, J. K. Vanos, and N. A. Kenny, 2016: The relationship between neighbourhood tree canopy cover and heat-related ambulance calls during extreme heat events in Toronto, Canada. Urban For. Urban Greening, 20, 180–186, https://doi.org/10.1016/j.ufug.2016.08.005.
Greaves, B., R. Trafford, N. Driedger, R. Paterson, D. Sills, D. Hudak, and N. Donaldson, 2001: The AURORA nowcasting platform—Extending the concept of a modifiable database for short range forecasting. Preprints, 17th Int. Conf. on Interactive Information and Processing Systems for Meteorology, Oceanography, and Hydrology, Albuquerque, NM, Amer. Meteor. Soc., 236–239.
Grimmond, C. S., G. Carmichael, H. Lean, A. Baklanov, S. Leroyer, V. Masson, K. H. Schluenzen, and B. Golding, 2015: Urban-scale environmental prediction systems. Seamless Prediction of the Earth System: From Minutes to Months, G. Brunet, S. Jones, and P. Ruti, Eds., World Meteorological Organization, 347–370.
Gultepe, I., and Coauthors, 2014a: Ice fog in Arctic during FRAM-IF project: Aviation and nowcasting applications. Bull. Amer. Meteor. Soc., 95, 211–226, https://doi.org/10.1175/BAMS-D-11-00071.1.
Gultepe, I., G. A. Isaac, P. Joe, P. Kucera, J. Thériault, and T. Fisico, 2014b: Roundhouse (RND) mountain top research site: Measurements and uncertainties for winter alpine weather conditions. J. Pure Appl. Geophys., 171, 59–85, https://doi.org/10.1007/s00024-012-0582-5.
Gultepe, I., and Coauthors, 2015: A review on ice fog measurements and modelling. Atmos. Res., 151, 2–19, https://doi.org/10.1016/j.atmosres.2014.04.014.
Gultepe, I., A. J. Heymsfield, P. Field, and D. Axisa, 2017: Ice phase precipitation. Ice Formation and Evolution in Clouds and Precipitation: Measurement and Modelling Challenges, Meteor. Monogr., Vol. 58, Amer. Meteor. Soc., 6.1–6.36, https://doi.org/10.1175/AMSMONOGRAPHS-D-16-0013.1.
Hayden, K. L., and Coauthors, 2011: Aircraft study of the impact of lake-breeze circulations on trace gases and particles during BAQS-Met 2007. Atmos. Chem. Phys., 11, 10 173–10 192, https://doi.org/10.5194/acp-11-10173-2011.
He, H., V. E. Fioletov, D. W. Tarasick, T. W. Mathews, and C. Long, 2013: Validation of Environment Canada and NOAA UV index forecasts with brewer measurements from Canada. J. Appl. Meteor. Climatol., 52, 1477–1489, https://doi.org/10.1175/JAMC-D-12-0286.1.
Herdt, A. J., 2017: A multi-index investigation of the spatiotemporal relationships between heat and EMS calls during the 2015 Pan American Games in Toronto, Canada. M.S. thesis, Dept. of Atmospheric Science, Texas Tech University, 87 pp.
Huang, L. X., 2016: Evaluation of nowcasts from integrated weighted model, integrated nowcasting system and mesoscale numerical weather forecasts for Environment Canada Pan Am Science Showcase. ECCC Internal Rep., 43 pp.
Huang, L. X., G. A. Isaac, and G. Sheng, 2012: Integrating NWP forecasts and observation data to improve nowcasting accuracy. Wea. Forecasting, 27, 938–953, https://doi.org/10.1175/WAF-D-11-00125.1.
Isaac, G. A., and Coauthors, 2014a: The Canadian Airport Nowcasting System (CAN-Now). Meteor. Appl., 21, 30–49, https://doi.org/10.1002/met.1342.
Isaac, G. A., and Coauthors, 2014b: Science of Nowcasting Olympic Weather for Vancouver 2010 (SNOW-V10): A World Weather Research Programme project. Pure Appl. Geophys., 171, 1–24, https://doi.org/10.1007/s00024-012-0579-0.
Jeong, C.-H., and Coauthors, 2010: Particle formation and growth at five rural and urban sites. Atmos. Chem. Phys., 10, 7979–7995, https://doi.org/10.5194/acp-10-7979-2010.
Jeong, C.-H., G. J. Evans, R. M. Healy, P. Jadidian, J. Wentzell, J. Liggio, and J. R. Brook, 2015 : Rapid physical and chemical transformation of traffic-related atmospheric particles near a highway. Atmos. Pollut. Res., 6, 662–672, https://doi.org/10.5094/APR.2015.075.
Jerrett, M., and Coauthors, 2009: A cohort study of traffic-related air pollution and mortality in Toronto, Canada. Environ. Health Perspect., 117, 772–777, https://doi.org/10.1289/ehp.11533.
Joe, P., and Coauthors, 2010: Weather services, science advances and the Vancouver 2010 Olympic and Paralympic winter games. Bull. Amer. Meteor. Soc., 91, 31–36, https://doi.org/10.1175/2009BAMS2998.1.
Johnston, D., and Coauthors, 2017: The Toronto 2015 Pan and Parapan American Games experience: An Environment and Climate Change Canada perspective. ECCC Rep., 81 pp.,http://donnees.ec.gc.ca/data/weather/predict/TO2015_Pan_and_Parapan_American_Games_Experience/TO2015_Pan_and_Parapan_American_Games_Experience.pdf.
Keenan, T., and Coauthors, 2003: The Sydney 2000 World Weather Research Programme Forecast Demonstration Project: Overview and current status. Bull. Amer. Meteor. Soc., 84, 1041–1054, https://doi.org/10.1175/BAMS-84-8-1041.
Kiktev, D., and Coauthors, 2017: FROST-2014: The Sochi Winter Olympics International Project. Bull. Amer. Meteor. Soc., 98, 1908–1929, https://doi.org/10.1175/BAMS-D-15-00307.1.
King, P. W. S., M. Leduc, D. M. L. Sills, N. R. Donaldson, D. R. Hudak, P. I. Joe, and B. P. Murphy, 2003: Lake breezes in Southern Ontario and their relation to tornado climatology. Wea. Forecasting, 18, 795–807, https://doi.org/10.1175/1520-0434(2003)018<0795:LBISOA>2.0.CO;2.
Klaassen, J., 2016: Summer 2012 Compact Weather Station Sensor Intercomparison Study, in support of the Pan Am and Parapan Am Games. ECCC Internal Rep., 188 pp.
Leroyer, S., S. Bélair, S. Z. Husain, and J. Mailhot, 2014: Subkilometer numerical weather prediction in an urban coastal area: A case study over the Vancouver metropolitan area. J. Appl. Meteor. Climatol., 53, 1433–1453, https://doi.org/10.1175/JAMC-D-13-0202.1.
Leroyer, S., S. Bélair, L. Spacek, and I. Gultepe, 2018: Modelling of radiation-based thermal stress indicators for urban numerical weather prediction. Urban Climate, in press.
Levy, I., and Coauthors, 2010: Unraveling the complex local-scale flows influencing ozone patterns in the southern Great Lakes of North America. Atmos. Chem. Phys., 10, 10 895–10 915, https://doi.org/10.5194/acp-10-10895-2010.
Levy, I., C. Mihele, G. Lu, J. Narayan, N. Hilker, and J. R. Brook, 2014: Elucidating multipollutant exposure across a complex metropolitan area by systematic deployment of a mobile laboratory. Atmos. Chem. Phys., 14, 7173–7193, https://doi.org/10.5194/acp-14-7173-2014.
Li, J., and H. W. Barker, 2005: A radiation algorithm with correlated-k distribution. Part I: Local thermal equilibrium. J. Atmos. Sci., 62, 286–309, https://doi.org/10.1175/JAS-3396.1.
Makar, P. A., and Coauthors, 2015: Feedbacks between air pollution and weather, part 2: Effects on chemistry. Atmos. Environ., 115, 499–526, https://doi.org/10.1016/j.atmosenv.2014.10.021.
Mariani, Z., A. Dehghan, P. Joe, and D. M. Sills, 2018: Observations of lake breeze events during the Toronto 2015 Pan-American Games. Bound.-Layer Meteor., 166, 113–135, https://doi.org/10.1007/s10546-017-0289-3.
Masson, V., 2000: A physically-based scheme for the urban energy budget in atmospheric models. Bound.-Layer Meteor., 94, 357–397, https://doi.org/10.1023/A:1002463829265.
Milbrandt, J., S. Belair, M. Faucher, M. Vallee, M. Carrera, and A. Glazer, 2016: The pan-Canadian High Resolution (2.5-km) Deterministic Prediction System. Wea. Forecasting, 31, 1791–1816, https://doi.org/10.1175/WAF-D-16-0035.1.
Moussa, S. G., and Coauthors, 2016: Emissions of hydrogen cyanide from on-road gasoline and diesel vehicles. Atmos. Environ., 131, 185–195, https://doi.org/10.1016/j.atmosenv.2016.01.050.
NAPS, 2017: NAPS data products. Accessed 15 Mar 2015, http://maps-cartes.ec.gc.ca/rnspa-naps/data.aspx?lang=en.
Oke, T. R., 2008: Urban observations. Guide to Meteorological Instruments and Methods of Observation: Part II. Observing Systems, 7th ed. World Meteorological Organization WMO 8, II-11-1–II-11-25.
Prüss-Ustün, A., J. Wolf, C. Corvalaìn, R. Bos, and M. Neira, 2016: Preventing disease through healthy environments: A global assessment of the burden of disease from environmental risks. WHO Rep., 150 pp.
Purdom, J. F. W., 1976: Some uses of high resolution GOES imagery in the mesoscale forecasting of convection and its behavior. Mon. Wea. Rev., 104, 1474–1483, https://doi.org/10.1175/1520-0493(1976)104<1474:SUOHRG>2.0.CO;2.
Rison, W., R. J. Thomas, P. R. Krehbiel, T. Hamlin, and J. Harlin, 1999: A GPS-based three-dimensional lightning mapping system: Initial observations in central New Mexico. Geophys. Res. Lett., 26, 3573–3576, https://doi.org/10.1029/1999GL010856.
Robichaud, A., and R. Ménard, 2014: Multi-year objective analyses of warm season ground-level ozone and PM2.5 over North America using real-time observations and Canadian operational AQ models. Atmos. Chem. Phys., 14, 1769–1800, https://doi.org/10.5194/acp-14-1769-2014.
Sills, D. M. L., 2009: On the MSC forecasters forums and the future role of the human forecaster. Bull. Amer. Meteor. Soc., 90, 619–627, https://doi.org/10.1175/2008BAMS2657.1.
Sills, D. M. L., and N. M. Taylor, 2008: The Research Support Desk (RSD) initiative at Environment Canada: Linking severe weather researchers and forecasters in a real-time operational setting. 24th Conf. on Severe Local Storms, Savannah, GA, Amer. Meteor. Soc., 9A.1, https://ams.confex.com/ams/24SLS/techprogram/paper_142033.htm.
Sills, D. M. L., P. Taylor, P. King, W. Hocking, and I. Nichols, 2002: ELBOW 2001—Studying the relationship between lake breezes and severe weather: Project overview and preliminary results. Preprints, 21st Severe Local Storms Conf., San Antonio, TX, Amer. Meteor. Soc., 611–614.
Sills, D. M. L., N. Driedger, B. Greaves, E. Hung, and R. Paterson, 2009: iCAST: A prototype thunderstorm nowcasting system focused on optimization of the human-machine mix. Proc. Second World Weather Research Programme Symp. on Nowcasting and Very Short Range Forecasting, Whistler, BC, Canada, Environment Canada, 2.9.
Sills, D. M. L., J. R. Brook, I. Levy, P. A. Makar, J. Zhang, and P. A. Taylor, 2011: Lake breezes in the southern Great Lakes region and their influence during BAQS-Met 2007. Atmos. Chem. Phys., 11, 7955–7973, https://doi.org/10.5194/acp-11-7955-2011.
Sills, D. M. L., H. Yang and P. Joe, 2014: A lightning mapping array in southern Ontario, Canada, and uses for severe weather nowcasting. 27th Conf. on Severe Local Storms, San Antonio, TX, Amer. Meteor. Soc., 83, https://ams.confex.com/ams/27SLS/webprogram/Paper255000.html.
Smith, K. R., and Coauthors, 2014: Human health: Impacts, adaptation, and co-benefits. Climate Change 2014: Impacts, Adaptation, and Vulnerability. Part A: Global and Sectoral Aspects, C. B. Field et al., Eds., Cambridge University Press, 709–754.
Stieb, D. M., R. T. Burnett, M. Smith-Doiron, O. Brion, H. H. Shin, and V. Economou, 2008: A new multipollutant, no-threshold air quality health index based on short-term associations observed in daily time-series analyses. J. Air Waste Manage. Assoc., 58, 435–450, www.tandfonline.com/doi/pdf/10.3155/1047-3289.58.3.435?needAccess=true.
Stroud, C. A., and Coauthors, 2011: Impact of model grid spacing on regional- and urban-scale AQ predictions of organic aerosol. Atmos. Chem. Phys., 11, 3107–3118, https://doi.org/10.5194/acp-11-3107-2011.