Boston Example Problems

Try out these examples to learn the basics of modeling oil spills in Boston Harbor and vicinity. Explore how changing tides, winds, runoff from the Merrimack River, and wastewater outflow can affect the trajectories of oil slicks. In addition, one example demonstrates how model and observation limitations can be overcome by including Uncertainty in the model solution. This knowledge will help you in designing your own WebGNOME model runs.

Example 1

Tides are an important part of the circulation in Boston Harbor. In this example, you will examine the effects of tides by starting a spill at two different times in the tidal cycle, once at the beginning of a flood tide:

Start Time: May 4, 2000 at 6:00 p.m. (1800)

and once at the beginning of an ebb tide:

Start Time: May 5, 2000 at midnight (0000)

The winds at the time of the spill are 5 knots from the SE. In this example, we won’t simulate the effects from the Effluent Outfall Tunnel. We will place an approximately 100-barrel spill near the Deer Island entrance to Boston Harbor.

Begin by selecting the Boston and vicinity Location File which will launch the Wizard to guide you through setting up the scenario. Use the information in the following table as you advance through the Wizard.

Start time:	As above for flood or ebb tide.
Model duration:	1 day.
Uncertainty:	Not included.
Sewage outfall effects:	Don’t consider outfall.
Wind:	5 knots from the SE.
Spill type:	Instantaneous.
Time of Release:	Same as model start time.
Amount released:	100 bbls.
Pollutant type:	Non-weathering.
Position:	42° 20.45’ N, 70° 57.44’ W.

Once you have run the model for both start times, zoom in to the spill area and examine the differences in beach impacts between the two spills. To more easily visualize the difference between the spill impacts, you could take a screenshot of the map at the end of the first run.

How do the two spills differ in the amount and location of pollutant?

Tip:

1. If you only want to change one spill parameter (like the start time of the spill in this example), there is no need to reload the Location File and step through the Wizard. Instead, use the buttons on the Menu Bar to switch from Map View to Setup View. In Setup View there are various panels which allow you to edit the model setup. In the Spill panel, click on the edit (pencil) icon of the spill you created to edit the time of release.

When you change the start time of the spill, you will likely want to change both the spill start time (done via clicking on the edit icon on the Spill panel) and the model start time (via the edit icon on the Model Settings panel). If you change the spill start time first, WebGNOME will automatically prompt you to change the model start time to match the spill start time. So it is a good idea to always change the spill start time first.

2. In the Map View, you can open the Layers panel (top right) to select different basemaps for familiarizing yourself with the local geography.

Answer:

When the spill starts just before the flood tide, oil reaches the Charles River and the northern portion of Boston Harbor. The oil beaches in these areas and is not immediately carried back out with the ebb tide. When the spill starts at the beginning of the ebb tide, the oil also impacts the western portion of Broad Sound.

Example 2

Wind can have a significant effect on a spill because it both moves the oil along the water’s surface and drives currents. Rerun the ebb tide spill (May 5 at 0000) with 5-knot wind from the NW, then with no wind.

How do the oil’s trajectory and shoreline impacts change from the previous example?

Tip:

To change the wind conditions in WebGNOME, in Setup View, click on the edit icon in the Wind panel then enter the wind speed and direction in the Constant Wind tab. Because you are entering a “Constant Wind” the Date & Time Field is not important as the same values will be applied at all model times.

For the “no wind” case, you either set the wind speed to 0 or delete the Wind from the list.

Answer:

Even a very light wind dramatically changes the oil’s trajectory! With the SE wind in Example 1, the spill moved to the northwestern regions of the harbor. Now, with the NW wind, the spill moves to the southeast, oiling the northern shores of Long Island, Houghs Neck, Peddocks, Hull, and other harbor islands. Without the wind, the oil does not spread as far north or as far south as it does with the wind’s assistance.

Example 3

Forecasts of environmental parameters are inherently uncertain. For example, wind and weather forecasts can be “off” in the speed, direction, or timing of winds. WebGNOME supports an “Uncertainty” solution in addition to the “Best Estimate” solution that you have been running. The “Uncertainty” solution takes into account our uncertainty in wind, horizontal mixing, and currents.

Rerun the previous scenario, increasing the wind to 10 knots from the NW. This time, run WebGNOME with the “Uncertainty” solution turned on.

Examine the difference between the “Best Estimate” (black) and “Uncertainty” (red) trajectories. Why do you think this type of information would be useful?

Tip:

To include the “Uncertainty” solution, click the box labeled “Include uncertainy in particle transport” in the Model Settings panel in Setup View.

Answer:

The “Uncertainty” solution shows where else the spill could go if the currents, winds, or other model inputs were a little bit different. In this case, the “Uncertainty” solution shows that the spill impacts could be more severe in the northern and western regions of the harbor, with the possibility of pollutant reaching areas near Deer Island, Spectacle Island, and Boston Inner Harbor. To the east, the spill could also be more extensive, with oil floating north of Little Harbor and Cohasset Harbor.

Responders use both the “Best Estimate” and “Uncertainty” trajectories to make decisions about how they will allocate response resources. Sometimes a highly valued environmental resource (e.g. an endangered species) may be important enough to protect, even if it has a low probability of being oiled.

Further Examples

The next few examples are brand new scenarios. You can choose to edit the various components as we did in the previous examples, or you may find it easier to re-load the Location file and step through the Wizard (just choose “Select a Location File” from the New pull down menu).

For these examples, turn off the “Uncertainty” solution and don’t include effects from the Sewage Outfall (we’ll learn how to turn this on in Example 6.)

Example 4

The Merrimack River has very high flows in the spring. This strong pulse of fresh water into the Gulf of Maine leads to a coastal current in Massachusetts Bay. Run two 1-day spill scenarios of 70,000 gallons of non-weathering oil near the entrance to Gloucester Harbor at 42° 34.73’ N, 70° 38.97’ W. Run one scenario during the spring freshet, on May 15, 2000 at 3:45 p.m. Run another scenario during the fall on October 15, 2000 at 7:15 p.m. Both of these times represent the beginning of a flood tide. In each case, there are no winds.

How does the oil’s trajectory change from the spring to the fall example?

Tip:

Remember to change the spill release time first rather than the model start time to automatically synchronize these two.

Answer:

In the spring, much of the oil is pushed to the southwest, away from Gloucester Harbor. In the fall, however, most of the oil ends up in the harbor.

Example 5

Different types of pollutants weather differently. In the previous examples, the pollutant that spilled did not change with time (it was “non-weathering”). Now you are going to run a 1-day scenario that compares the effects of different types of pollutants.

A damaged vessel begins to leak fuel as it heads into Boston Harbor along Nantasket Roads. The vessel spills 30,000 gallons of product between 6:00 and 7:00 p.m. on May 4, 2000 as it travels from 42° 19.16’ N, 70° 53.55’ W to 42° 18.76’ N, 70° 55.25’ W. Winds are 5 knots from the W.

A scenario that includes weathering requires additional water property information. In the Water panel, enter a water temperature of 55° F and use the defaults for the other parameters.

Run the above scenario for a barge carrying a heavy fuel oil (e.g., an IFO or Intermediate Fuel Oil) and for a fishing vessel carrying a diesel fuel. At the end of your 24-hour prediction, write down the mass balance for each product in the table below.

	IFO (gallons)	Diesel (gallons)
Released	30,000	30,000
Floating
Beached
Evaporated
Dispersed

Tips:

1. The spills you have set in previous examples have been point source spills, representing catastrophic releases at a particular time and location. The spill described in this example is a spill from a vessel that is leaking as it is moving. To model this “line source” spill in WebGNOME, you’ll create a Continuous Release. If you are using the Wizard, you’ll simply choose this option. If you are modifying the scenario using Setup View, start by deleting existing spills from the previous examples. Then click on the Create Spill Icon (plus sign) in the Spill panel. Choose Continuous Release.

The spill duration in this example is 1 hour. Use the Add Endpoint button to specify a spill along a transit.

2. Use the ADIOS Oil Database link to open the ADIOS oil database. From the database interface you can select an oil that corresponds to an IFO or a diesel fuel. It doesn’t matter which exact oil you select, as long as it falls into one of these broad categories. Download the oil and load the file into WebGNOME using the load oil drop box.

3. To view the mass balance for each scenario switch to the Fate View.

Answer:

Heavier oils remain in the environment longer than lighter, refined products. You should see that much more diesel has evaporated and dispersed than fuel oil #6 after 24 hours in the water.

Example 6

The new Effluent Outfall Tunnel discharges wastewater in Massachusetts Bay, about 14 km from Boston Harbor. Numerical modeling studies suggest that the discharge will have little or no effect on surface currents in the region in the summer and a small effect on surface currents in the winter. To demonstrate how this discharge may impact spilled oil trajectories, run a 100-barrel linear spill over the outfall site from 42° 24.97’ N, 70° 47.04’ W to 42° 21.96’ N, 70° 46.98’ W. Leave the substance as the diesel used in the last example. Start the spill at noon on February 15. Run WebGNOME twice for this spill, once with the sewage outfall effects option turned on, and once with no sewage outfall effects.

Model Parameters:

Start time:	February 15, 2000 12:00.
Model duration:	1 day.
Uncertainty:	Not included.
Sewage outfall effects:	Run both cases.
Wind:	5 knots from the SE.
Spill type:	Instantaneous.
Time of Release:	Same as model start time.
Amount released:	100 bbls.
Pollutant type:	Non-weathering.
Position:	42° 24.97’ N, 70° 47.04’ W to 42° 21.96’ N, 70° 46.98’W.

How does the wastewater outfall impact the oil’s trajectory?

Tip:

To include the sewage outfall effects without relaunching the Wizard, you will change a setting in the Current panel. Within this panel is a list of surface current patterns that are described in the Location File User Guide. Find the pattern labeled “Sewage Outfall Current”. The checkbox next to the current name is unchecked if you chose not to consider the outfall effects when you set up the scenario. Click the checkbox to include this current pattern.

Answer:

Few, if any, discernible changes result from adding the effects of the sewage outfall; however, in the scenario that includes the sewage outfall effects, there is an area in the middle of the trajectory that tends to remain clear of oil.