Sunshine, Taxes, and Maps

I’ve recently been looking at economic incentives for solar power in the U.S.  I thought I’d consider what incentives might encourage the installation of solar panels for different locations across the country.

First, let’s see how much sunshine different parts of the country get annually:

Click for better resolution
This plot is generated in R using the shapefile GIS data provided by the National Renewable Energy Laboratory (NREL) via the Open Energy Information (OpenEI) platform.  It is based off a model that takes input from 14,000+ solar radiation stations across the country.

Naturally, increased sunlight can provide more incentive to install solar panels.  However, that is only a small part of the story.

What happens when we recreate the above map, but measure the economic value of installing the solar panels due to:

Clearly, there’s quite a bit of information to coalesce.  To simplify things, I consider two scenarios:

The Residential Scenario:

Source: http://256.com/solar/ Used with permission
  • 20 sq. meters of solar panelling installed on a single family residence, unobstructed
  • Solar panels show a 10% energy conversion efficiency (reasonable, if low)
  • Installation costs $50,000
  • Home owners bear a tax rate of approximately 15%

The resulting gross payout of tax incentives and energy production (not subtracting the cost of installation on the residence) extrapolated over time, looks as follows:  (Be sure to click for the high resolution version)

Click for better resolution


In addition to the residential scenario above, I also considered a more commercial scenario, which involved a different set of state level tax incentives, and a few other assumptions.

The Commercial Scenario:

Source: http://schools-wikipedia.org/images/258/25899.jpg.htm Used with permission
  • 200 sq. meters of solar panelling installed on a business, unobstructed
  • Solar panels show a 10% energy conversion efficiency
  • Installation costs $500,000
  • Corporation bears a tax rate of approximately 30%

The resulting gross payout of tax incentives and energy production (not subtracting the cost of installation on the residence) extrapolated over time, looks as follows:  (Be sure to click for the high resolution version)

Click for better resolution

 

Naturally, there are a huge number of assumptions baked into these maps.  (Stable tax rates, typical weather, steady energy prices … the list goes on.)  Nevertheless, I think it’s interesting to see, at least according to the data I used, how tax write-offs start out dominating the state-to-state variation in payout.  But, after a few decades, solar potential begins playing a more critical role.

 

A few things I learned along the way:

  • I taught myself quite a bit about R in this little study.
  • Using industry standard tools for GIS for the first time (namely becoming familiar with the structure of a “shapefile”).  This was probably the most challenging part.
  • In dealing with tax incentives, the number of stipulations and caveats can be absolutely crushing.  I found the best way to deal with these were to simply hypothesize a couple of scenarios, and run with the results.

Questions and comments are welcome.

If Madison Crime were Elevation

What if Madison, WI were mapped so that elevation represented crime?  This idea is directly inspired by Doug McCune’s post about mapping San Francisco crime as elevation from last year.

Accessing the data was a little tricky.  In Madison, the police put incident reports online, but some scrubbing of the data is required, and the reports needed to be geocoded.  After a little scripting to overcome these obstacles, I was ready to start mapping.

I divided crimes into 3 broad categories:

  • Robbery/Burglary/Theft
  • Drugs and Alcohol
  • Violent Crime

Here are the results:

Noticeable features:

  • The State Street area (at map-center), linking the capitol and the UW-Madison campus, has the largest spike of robbery activity.  (This will be a recurring theme…)
  • The shopping centers, in particular the malls, have significant numbers of reports.

 

 

Noticeable features:

  • State Street is again an unpleasant place to be.
  • The south side (Park St / Fish Hatchery Rd) has its fair share of violent crime, as does the shopping center where Verona Rd meets the Beltline Hwy.
  • Madison’s east side is pretty chilled out.  Perhaps this correlates with the better pizza options?

 

 

Noticeable features:

  • Not nearly so much trouble with State St
  • The biggest trouble in town seems to be on the east side.

 

A few things I learned along the way:

  • How to build a crude web-spider using cURL.
  • How to geocode addresses using the Google Geocoding API — remarkably user-friendly!
  • A few basics about MySQL to store the scraped and geocoded data.
  • Quite a bit about a ray-tracing program called POV-Ray.
  • How to animate graphics using imagemagick — though, the resulting animations (showing the distribution of crimes changing along with the time of day) are far too big to put on this site.

Please leave any questions or comments.