Introducing Purple Roof's Green Roof Modeler

What is green roof hydrologic modeling?

Modeling is becoming increasingly relevant in design and construction. But modeling might sound like a scary proposition. Is it hard? Do I need to learn new skills? Is it time-consuming? And do I need to model anyway? We hope this article convincingly answers: “No, No, No, and Yes” to those questions.

Let’s dig in.

Go straight to Purple-Roof Green Roof Modeler

Purple-Roof DETAILS

First, what do we mean by model?

A model, in the sense we are using the term, is a mathematical representation of predicted performance. This is similar to a weather model, predicting a hurricane, for example.

There are inputs, such as current storm size, wind conditions, sea temperatures, and there are outputs, such as the trajectory of the storm, wind speed at landfall, and anticipated rainfall volume once the hurricane reaches shore. A model that predicts performance over multiple, sequential timesteps is a continuous flow model. We use the term modeler for the program that generates the model.

Modeling probably has a reputation for being difficult. An architect who wants to model a building using BIM might have had a steep learning curve from AutoCAD. And a civil engineer or hydrologist might consider it difficult to learn a program such as HEC-RAS.

Purple-Roof’s Green Roof Modeler is designed to be easy! We want anyone, regardless of education, experience, or skill level to be able to use the modeler. We think it’s very important for green roof performance concepts to be easily understood and universally available.

The difficulty in learning or using most other modeling software often lies in the extensive input required. There are pros and cons. Pros of having a granular level of input is that you have greater control over the model’s output. Cons are that greater user control can lead to errors and that this makes the program seem daunting, so fewer people use it.

We have chosen to use the absolute minimum user inputs for our first version of the modeler, and later versions will maintain the absolute simplest interface possible but will layer in options for greater user control.

Step 1: we want people to see the value of modeling green roofs!

The rest of this post is a “How-To.” Check out this post on the benefits of modeling green roofs. There is also some information about the modeling project on the modeling home page, and all the current modeling equations are at the bottom of the modeling home page.

Generating models

Enter a US location and a year. This will search weather databases for historical weather for that location for that year. The program will return an error if the data is not found. In later versions, we will add an option for stochastic weather by region, which we believe will be more accurate and faster. Stochastic weather is statistically accurate weather for an area, but which does not represent any particular timeframe.

Reading models

The model is presented as a simple bar graph. The horizontal axis is time, from the first date on the left to the last date on the right, usually January 1 to December 31.

The top part of the graph is weather. Weather data is obtained from a 3rd party weather service. Three weather data points shown are humidity, temperature, and precipitation. Humidity is the lower horizontal bar, ranging from yellow (low humidity) to green (high humidity). Temperature is the upper horizontal bar, ranging from red (hot) to blue (cold). Precipitation bars are blue, variable-length vertical bars at the top of the screen. The y-axis (vertical axis) for precipitation is labeled on the left, ranging from no precipitation to 100mm (2 inches) of precipitation at the top of the scale.

Hover over any of the weather or performance bars to see numerical information for that date at the top of the screen.

Weather data includes many more data points than are currently shown, such as wind and solar intensity.

The bottom part of the graph is performance. Two performance metrics, plant stress, and runoff are shown in variable-length vertical bars. Plant stress is shown in red from no stress (no bar) to low stress (short bar) to very high stress (long bar). Runoff in purple is basically the mirror image of precipitation ranging from no runoff to 100mm (2 inches) of runoff at the bottom of the scale.

You may toggle units between SI and imperial. If you accepted the user experience cookie, your browser will save your choice.

At the very bottom of the model, there are three dashboard graphs: total volume of precipitation and how much of that precipitation leaves the roof via evapotranspiration vs. runoff.

Go straight to the Purple-Roof Green Roof Modeler

Precipitation only leaves the roof in these two ways (gravitational runoff or water vapor). Evapotranspiration and water retention are the same value, as all water retained eventually leaves as water vapor.

Total days at different stress levels. Elimination of all stress might not be a reasonable goal, but we believe a minimization of stress should be a goal. This dashboard summarizes stress data shown on the overall graph.

Runoff days by volume. This quantifies occurrences of runoff at different thresholds. This only considers the number of days rain occurred. Since this version of the modeler does not consider detention, runoff is assumed to occur instantaneously, so days with no rainfall also have no runoff.

Comparing Models

Now that you are oriented to reading the graphs, let’s change the graphs. The current version of the Green Roof Modeler generates 40 models and lets you toggle between them and compare each against a baseline. Those 40 models are different combinations of soil depth (50mm [2 inches] to 150mm [6 inches]), mineral wool depth (none to 75mm [3 inches]), and detention layer (not included, or included).

The model first loads showing a baseline of 100mm (4 inches) of green roof soil, with no mineral wool and no detention layer. This baseline was selected as it is one of the most common green roof profiles. You may slide any of the toggles to change the green roof profile. If the profile lowers retention or lowers stress versus the baseline, the delta between baseline and selected profile will be visible. The baseline is always in the background in a faded tone.

So what is this good for?

The current version of the Green Roof Modeler is intended to illustrate comparative performance differences between profiles. We think this is a first step toward taking a more intelligent approach to designing and specifying green roofs. In the very near future we will allow you to download the model as a CSV file, meanwhile contact us if you would like to download a particular model.

Purple-Roof’s Green Roof Modeler is an ongoing project that has just begun. We want your feedback! Drop us a line to let us know what improvements or features would benefit you. Let us know your questions or if anything is not clear. We want to develop this tool working with a community of hydrologists, civil engineers, architects, and landscape architects as part of a movement to improve green roof performance.

Go straight to Purple-Roof Green Roof Modeler

Purple-Roof HOME

Reading Tip - Hydrologic Modeling

She, Nian, and Joseph Pang. "Physically based green roof model." Journal of hydrologic engineering 15.6 (2009): 458-464.

Hydrologic Modeling Software