Friday, March 15, 2013

Answer: How much death at the roadside?

The synopsis: around 3.1 million pounds / year. But there are lots of details here... 

We sometimes drown in data, and this Challenge turns out to be a hard problem not because there’s too little data, but because there’s too MUCH data and because so much of it is so similar!

The Challenge asks “Just how many pounds of herbicide are sprayed along California roads each year?”  I limited the definition of “California roads” to just “state roads” largely because which organization maintains the roadsides varies tremendously depending on ownership.  Country roads are run (and sprayed) by the various counties; city roads likewise; but STATE roads are owned and maintained by the California Department of Transportation (aka CalTrans). 

I spent a good deal of my times trying to figure out what would be an authoritative source to answer this question.  I began with  the query:

     [caltrans roadside OR "road side" herbicide]

I used “Caltrans” because I knew that would be the preferred name of the “California Department of Transportation.”  I quickly learned that the official Caltrans web site is, but that the term Caltrans is used almost everywhere as a shorthand for the longer name. 

This query worked pretty well, but got a lot of anti-herbicide advocacy groups (e.g.,,, and  The reality is probably somewhere hidden in all of their writings, so this challenge mostly boils down to figuring out what the consensus of opinion is about roads, spraying, pounds of herbicide and kinds of chemicals.  So in a sense, this is a great problem to learn how to read highly charged content from multiple points of view. 

I decided to work from the Caltrans point-of-view first, hoping that they’d have fairly straightforward data about roads, herbicides, and patterns of spraying. 

It didn’t take me long to get to the 5th result, "Use of herbicides in roadside environments," which led me into the maze of state reports and sometimes confusing masses data. 

I spent around 2 hours looking at different Caltrans documents, trying to get to what looked like a decent data set about roadside spraying.

I learned a few things along the way… mostly about how the state of California is organized, but more importantly (for our purposes) what special language they use to encode their information.   

Adding to the confusion, different herbicides often have a large variety of names.  There are product names (Roundup; Eraser... vs. chemical names (e.g., glyphosate, Diammonium salt; glyphosate, dimethylamine salt; 2,4 D … ) 

One of the real treasures I found was that the California state website,, has a department of “Pesticide Regulation”  They have a database system “CalPIP” that provides the ability to drill down (e.g., by county, pesticide name, chemical name, or zip code). 

I also learned that the term-of-art used to describe roadside application is “rights-of-way.”  So when searching through data, that’s the common term used across all of California data. 

But what I wanted was a data sheet with pounds-applied.  Now, after 2 hours, I’d found several decent looking data sets. 

One of my good queries was for:

    [ herbicide California right-of-way ]      -- note that right-of-way also matches "right of way"

which led me to the site ( ) which aggregates data from Caltrans over the past several years.  You can see the “right of way”  (note the variation in the term!!!) data for1991 – 2009 here.  

From this data, I was able to make the following chart of active-component herbicide use on rights-of-way: 

This data looked about right, but they’re an advocacy group lobbying to restrict the overuse of pesticides.  So I wanted to check their data.  I did that by using the CalPIP site I mentioned above. 

From CalPIP I got a data table of ALL reported pesticide use in the state by Caltrans for 2010 (the latest data I could find) and imported it into my spreadsheet app.

I filtered by herbicide (excluding all of the other –cides), then filtered by “rights of way” uses (that is, roadside application only) then added up all 49,190 reported instances of herbicide uses on the sides of the road.  Answer?  3.1 million pounds.  Bear in mind that this is “active chemical” only—they guys on the trucks also mixed in other stuff (called “adjuvants”) such as diesel oil or other sticky / anti-spray-drift compounds.  They actually sprayed 8.3 million pounds on the roadside.

In any case, here’s the top of that spreadsheet:

and the relevant summation:

(The little green triangles in the upper left of each summary number are complaining that some of the cells have "N/A" in them.  This doesn't really affect the total numbers.)  

The good news here is that this number fits in very plausibly with the rest of the data from the chart above, so I’d think of this as confirming the data from (which, to their credit, links to the Caltrans website as their original data source, so I’m not terribly surprised). 

Search lesson:  There are many to point out here.  This was a big, complex task.  So let me list my take-aways. 

1.  Big tasks require taking notes.  I had a Google Doc open in a separate window on a second monitor.  I don’t know how I would have kept track of everything I was doing (and all of the various reports I read) without some systematic notes.  However…

2.  You can re-find data if you copy enough of the original into your notes.  When I was creating the chart above I at one point couldn’t remember where I’d found that data.  Luckily, I was able to put in a bunch of the data (suitably double-quoted) and was able to re-discover where I’d gotten it from. 

3.  Be aware of “terms of art” in your reading.  The example of “rights of way” was important.  I’d tried many searches with variations on “roadside spraying” or “highway herbicide”—but they were all too general.  Once I noticed “rights of way” as a specialist’s term, I was able to really focus in on what I wanted to find. 

4.  Look for multiple resources all pointing to the same place. The first few documents I read all pointed back to the site, and once I started reading around, I kept seeing a few resources that were being referred to a great deal.  That was a handy insight to have.

5.  There is diversity even in the government.  One tends to think of government sites as monolithic and all hewing to the “government’s story.”  But I was surprised by how much variation there was between different government web pages, even those that are telling more-or-less the same story.  Some documents discuss how non-toxic the herbicides are, others tell a different story.  Again, the truth is out there, you just have to collect the data and evaluate it for yourself. 

And... I'll get around to the relative toxicity of the herbicides used in another post.  It's a bit too much to do all in one day! 



I started out this whole process by wondering just how many miles of state roads CalTrans operates.  The obvious query: 

[ caltrans miles of roads ]

First result:  leads to  >50K miles of roads. 

But the second result: says that the entire Interstate Highway System is 46.5K miles!  Does the state of California have more miles of roads than the entire US Interstate System? 

Looking at a few other sites, I found a few references claiming that California has only ~15K miles of roads. 

What’s going on here?

After looking through a few more pages, I landed on the CalTrans document listing total road miles and LANE miles.  

There, Table 1 points out that CalTrans has 50.46K miles of LANES and 15.1K miles of ROADS.  Ah ha!  (That same table points out that the Golden Gate Bridge Authority has 15.8 miles of lanes but only 3.38 miles of roads.  Since I’ve driven all those roads, I know they’re all 4 lanes wide (at least), so now we see where the two numbers come from:  LANES vs. ROADS.) 

6.  Be sure you understand the basis on which your claims are based.  Although the difference between 50.46K and 15.1K seems huge, the truth is that it’s really the same number… just with different basis at work. 


  1. Dr. Russell. Great answer!

    "terms of art" are many times hard to find. I thougt as RoseMary that Right of Way encompassed much more than just roadsides until found that was the right name. I like to take notes also because if when posting my answer doesn´t show I can post it again.

    Finally, about how toxic is this herbicid? Found that worker can sue over herbicides. Also found that in California a year later some sprayed roads are without vegetation and even so, they want to spray again.

    And, today I just read an article that I like to share because it relates with toxicity. It is about pesticides and bees in EU: "Bee deaths: EU delays action on pesticides ban"

    Have a great week.

  2. Dr. Rick Relyea shows definitive research that the pesticides are killing amphibians.
    He also shows great courage in going up against the Darth Vader of companies.

    I live in the Sierra Nevada foothills where the brown roadside stripes go for miles and miles along major thoroughfares including in, over, and through water. How did this happen? This is California and we are dumping thousands and thousands of pounds of herbicides onto our state every year now, sometimes multiple sprayings. Was a CEQA document prepared for this? It is no doubt a blast of poison putting the last nails in the coffin of our roadside amphibian populations. This has to stop.

  3. I learned alot from this challenge. I had a preconceived idea about right of way because of my background and for me ROWs included so much more. Narrowing down to Caltrans should have been my clue. So terminology has to apply to the subject and itmay have a regional tone to it as well.
    Secondly I am new to online databases and when I was not getting what I wanted I spent time trying to understand how to use this tool. Working from a tablet can be restrictive but I still could have gotten figures. I would appreciate advice on making the best use of the data. I saw another student got all the data (well done!) but it didn't zero in on the specific data we needed. Any tips on how these work generally, how to make our own tables and charts from them, and how to get just what we need would be very helpful. Links to online resources would be useful. I am hoping that once you have the basic concepts then it is just a matter of identifying info within a specific database.

  4. Wanting to learn more I went through the exercise as you described and stumbling a bit along the way. First I realize I have to use my desktop when using a database so if others our trying on a tablet you too may have difficulties. I downloaded both databases as described. I then imported them into a spreadsheet. I was able to create a chart (something I haven't done in many years) which looks a lot like yours. Then I downloaded the 2010 database. I selected the right of way filter and I got the same pounds of chemical applied exactly, my products applied were close. Now I filtered Products Applied leaving out all products for rodents, insects, adjuvants and other "obvious" products that weren't herbicides. I did it then realized you had actually accepted the N/A in the "Chemicals Applied" column so I could stop at this point. Is that why you accepted that conclusion?

    Sorry if this is too elementary for others ( I hope few others have questions) but I'm going from not knowing anything about using databases or importing data into spreadsheets. I see there are database applications and I downloaded a free one. Is this in fact a useful tool, more so than importing it into a spreadsheet? I haven't experimented with it yet. The above has been a sufficient challenge. Thanks.