I need some help interpreting the results from the valuation of the heat reduction service. I ran the data provided for the summer series workshop and I got the following results
Land Use Type Energy Saving
Residential Low Density 2 124.93
Commercial 1 148087.8
Commercial 1 28242.18
etc…
Does the energy saving column indicate to the kWh saved to cooling the building? In this case 148087.8 seems like a huge number!
Perhaps I am missing something. I will really appreciate any comments or feedback. Thanks!!
Laura
According to the User Guide, the “energy_sav” column within “buildings_with_stats[suffix].shp” provides savings for each building in total monetary units over your period of interest. However, unless you entered the optional input of energy cost (e.g.: $ / kWh) for each building type, the model would be unable to calculate such monetary savings. I suspect that you did NOT input energy costs, and therefore such results are not sensible. I actually have ‘correcting this information in the User Guide’ on my list of tasks to do, so I’m glad you are forcing us to prioritize that, but I apologize for the confusion it’s causing you in the meantime.
My understanding is that when energy costs are not provided as an input, the “energy_sav” column actually provides results for each building in total kWh saved, or mitigated, over the entire period of interest. (I believe that’s a YES to your question.) So, if your input data covered a single month (you stated in yesterday’s post that your analysis is of July in Phoenix, Arizona, USA), then those values represent the cumulative kWh saved over all of July. Most of the values you’ve shared seem reasonable to me, with the possible exception of those in the 100,000s range, like 148,087.8 and 219,075 for residential high density, especially for a single month. But, while those values are quite large, they may be reasonable given how HOT Phoenix’s mean temperature is in July (>40C). Did you use the cooling degree days approach? If so, to what temp did you assume buildings were being cooled (thermostat setting)?
Thank you for your clarification. Actually those numbers were from the dataset provided for the virtual workshop and refer to Minneapolis. I decided to take a step back and look at the example you guys presented to better understand the model.
My main doubt (and I apoligize for being so confused!) is that the values on the energy_sav column quantify the total amount payed (or the kwh used) to cool down the house and depend on the m2. You indicated that they represent “saving”…but saving compared to what?
I estimated the saving running alternative land use scenarios and comparing the kwh (or $) used. For instance, houses near a city park save more energy than houses in industrial zone. Again, I apologize for being so confused. I hope you can have some explanation as I am really going crazy trying to figure it out. By the way I am running the model to quantify the cooling capacity of urban lakes and it is the last chapter of my PhD!
There is no need to apologize. This is one of our newest InVEST models, thus we do not yet have as many real world use cases to draw from compared with our other models. We value feedback from users like yourself so that we can continually improve the models and their documentation. Thank you for being so active on the Forum! The way you’ve framed your questions reveals that the Urban Cooling chapter of the User Guide could likely benefit from some more precise language and clarifications.
For that “energy_sav” column, my understanding is that the values represent the energy savings provided by a city’s natural spaces, or green areas. So, the savings are compared to a theoretical version of the city that does NOT have ANY natural attributes within it’s land cover classes. The biophysical table assigns shade proportion, crop ET coefficient, and albedo to each landcover class. Each of these parameters contributes to a cooling capacity (CC) index for each class. So, the savings are compared to a scenario where all these values are equal to 0, as in no landcover class has any natural capacity to cool, or there are no natural spaces nor green areas in the entire city. (Conversely, building intensity values are used to calculate the cooling capacity index if using the night-time method, but in that case, savings would be compared to a scenario where that value = 1, meaning all land area is covered by a building footprint.) The heat mitigation index for a pixel is equal to the CC if it is unaffected by a green space >2 hectares, and adjusted by a distance-weighted average if it is affected. This is because large green spaces have a disproportionate cooling affect on their surroundings (see: Zardo et al., 2017 and McDonald et al. 2016 “Planting Healthy Air…”). So, the savings are also compared to a scenario withOUT any of these, or “Green_area” set to 0 for all classes.
That being said, the input files supplied in the Virtual Workshop data packet are for the Twin Cities in Minnesota as you said, but they contain ANNUAL averages. This means that the results also relate to an entire year, so the savings values you shared should seem somewhat more reasonable considering they are totals for 12 months, not just one month. However, I re-ran the model with these data to compare with your results and my numbers are considerably lower. For instance, my maximum savings for any single building is ~61,095 kWh. I notice that your last column, “mean_t_air” (average temp in building) has values >40C. These are way too hot for annual averages in Minnesota. All the values output in my run are <25C. I am curious how you parameterized the model. Did you refer to and enter the values given in the “README.txt” file that came with the data packet? Specifically:
Baseline air temperature (deg C): 23.2
Magnitude of the UHI effect (deg C): 2.05
Air Temperature Maximum Blending Distance (m): 600
Green Area Maximum Cooling Distance (m): 450
Average relative humidity (0-100%): 30
relative weight to apply to Shade: 0.6
relative weight to apply to Albedo: 0.2
relative weight to apply to Evapotranspiration Index: 0.2
Is the table you shared a screenshot of from results for the baseline “lulc.tif”, or is that from one of the alternate scenarios, perhaps “industrial.tif”?
I hope this makes sense and is helpful. Please let us know. We don’t want to contribute to craziness nor inhibit the completion of your PhD!
