Wednesday: 29 July 2009
The long range prediction for our area is for rain through next Tuesday. From the National Weather Service:
Oh please. “Glimmer of hope?” Give me a break.
Every week on Monday I herd cats. Actually I herd cats every day, but these particular cats are the Oglethorpe Countians who contribute to our rainfall measurements. Mostly I just peruse the CoCoRaHS data and then get the station totals. I put them into an excel file, then a text file, then copy and paste it to an email to our fine county weekly, the Oglethorpe Echo, where it gets published. Most of the time. There are three feral cats who don’t do computers and call in their totals for the week.
It’s sorta fun, and I’ve taken the opportunity to dip a lot more into the very nice resource of data that is the Community Collaborative Rain, Hail, and Snow Network, known by the ponderous, marginally retentive and inaccurately constructed acronym CoCoRaHS, which I get tired of typing and so will just call it CCR from this point on.
It’s not just the data - it’s also the various searches you can do on the network. The “view data” option lets you get totals for a county or state list of individual stations for any period of time, and the “maps” option shows you the totals for any particular data. There’s a lot more than just that.
So that’s the plug, but there’s a reason for it. I’ve taken advantage of that resource and capability, and have done a little analysis.
Oglethorpe County, of all places(!), has a number of reliable observers that matches the best in Georgia. This is certainly due to the organizational skills of the previous Oglethorpe County coordinator, who started us off in May 2008 with more observers than in any other Georgia county. For the most part, that’s remained stable.
There are enough observers to where we can pinpoint where we need more, in order to blanket the county. I’ve thrown a few of the colloquial district names onto this CCR map. You’ll recognize Wolfskin, of course, and it’s especially interesting that we have three or four stations very close together, within four or five miles. We could use some observers in southeast Oglethorpe, the Philomath area, and Vesta and Goosepond could also use some more. Beaverdam and Devils Pond are our most populous districts - surely there’s some talent there. Maxeys is a little difficult, since much is the very lightly populated area including the northern portion of the extremely fragmented Oconee National Forest.
I took a sampling, as of this date, of the stations in eight counties in Georgia. It’s no surprise that the densest coverage is around the big cities. If you know the geography you’ll be able to spot where Huntsville, Birmingham, and Montgomery, Alabama are.
Seven of the counties have the highest number of CCR stations, and I included Floyd County too, for sentimental reasons. Close to are the four counties that make up Atlanta, and Oglethorpe and Clarke Counties are Athens. Chatham County is Savannah. (I might note that of the four Atlanta counties, Cobb County is the home of the egregious Newt Gingrich. Our family lived in Dekalb County for three years in the late sixties. Gwinnett is so large that I shouldn’t generalize, but the parts you go through if you are so inclined to go to Atlanta are out of my worst nightmare.)
So here’s what I did, and you can suffer through the analysis or you can skip down to the bottom couple of paragraphs.
CCR gives the total number of stations - Oglethorpe, our very red county, is in the top four, even including Atlanta counties. But some stations are not reliable. CCR won’t count them as inactive unless the station operator says so, but in Oglethorpe, for instance, there are a couple of stations that gave a single report a year ago and then never again. CCR doesn’t drop any data, so those stations are still listed in the totals.
That’s why I have the column that I call “reliable.” It’s a judgement call, but if a station hasn’t reported in months, or if they make only intermittent reports when rain probably fell, if there are too many blanks when zeroes could have been reported, then I consider them unreliable. Oglethorpe County can usually count on 11 reliable reports, consistently, out of 19 stations.
The last three columns are all taken using July 2009 data, to date. I accepted CCR’s count of stations reporting Jul 1 to the present. The next to the last column uses that count as a rough estimate of the percent of reliable stations in the county. Mostly they hover aorund 60% of the total stations listed. Again, that’s mainly because CCR doesn’t delete stations. Oglethorpe maintains the highest percentage; Fulton County can only claim 40% of its total stations as being reliable. Our negative control, Floyd County, lost almost 50% of its stations to unreliability.
The last column is the percentage of stations reporting in July that I consider to be reliable. These generally run 80-90%, and Oglethorpe is right in there. But of the 61% reliable stations still in Dekalb County, 100% are reporting in July, and the same is true for Floyd County. Pretty impressive!
What to make of all this? In populous counties in Georgia, at least, only around 60% of CCR observers stick with it reliably, over the course of a year. Of those, 80-90% reporting in the last month are in the reliable group, leaving 10-20% of the total stations that occasionally give a report but can’t be counted on. I guess that would leave 20-40% who totally dropped out after a short time.
I don’t know how well CoCoRaHS represents volunteer organizations in general, but it fits some of the parameters. People are initially attracted by some point of interest. 20-40% drop out in the first year. Of the ones remaining, 10-20% are unreliable. In the end, you can count on about half the initial interested sticking with you, at least through the first year.
It looks fairly wet through the forecast period. The only glimmer of hope for drier weather is showing up in the 00z GFS which tries to build a high down through the Tennessee Valley early next week.
Oh please. “Glimmer of hope?” Give me a break.
Every week on Monday I herd cats. Actually I herd cats every day, but these particular cats are the Oglethorpe Countians who contribute to our rainfall measurements. Mostly I just peruse the CoCoRaHS data and then get the station totals. I put them into an excel file, then a text file, then copy and paste it to an email to our fine county weekly, the Oglethorpe Echo, where it gets published. Most of the time. There are three feral cats who don’t do computers and call in their totals for the week.
It’s sorta fun, and I’ve taken the opportunity to dip a lot more into the very nice resource of data that is the Community Collaborative Rain, Hail, and Snow Network, known by the ponderous, marginally retentive and inaccurately constructed acronym CoCoRaHS, which I get tired of typing and so will just call it CCR from this point on.
It’s not just the data - it’s also the various searches you can do on the network. The “view data” option lets you get totals for a county or state list of individual stations for any period of time, and the “maps” option shows you the totals for any particular data. There’s a lot more than just that.
So that’s the plug, but there’s a reason for it. I’ve taken advantage of that resource and capability, and have done a little analysis.
Oglethorpe County, of all places(!), has a number of reliable observers that matches the best in Georgia. This is certainly due to the organizational skills of the previous Oglethorpe County coordinator, who started us off in May 2008 with more observers than in any other Georgia county. For the most part, that’s remained stable.
There are enough observers to where we can pinpoint where we need more, in order to blanket the county. I’ve thrown a few of the colloquial district names onto this CCR map. You’ll recognize Wolfskin, of course, and it’s especially interesting that we have three or four stations very close together, within four or five miles. We could use some observers in southeast Oglethorpe, the Philomath area, and Vesta and Goosepond could also use some more. Beaverdam and Devils Pond are our most populous districts - surely there’s some talent there. Maxeys is a little difficult, since much is the very lightly populated area including the northern portion of the extremely fragmented Oconee National Forest.
I took a sampling, as of this date, of the stations in eight counties in Georgia. It’s no surprise that the densest coverage is around the big cities. If you know the geography you’ll be able to spot where Huntsville, Birmingham, and Montgomery, Alabama are.
Seven of the counties have the highest number of CCR stations, and I included Floyd County too, for sentimental reasons. Close to are the four counties that make up Atlanta, and Oglethorpe and Clarke Counties are Athens. Chatham County is Savannah. (I might note that of the four Atlanta counties, Cobb County is the home of the egregious Newt Gingrich. Our family lived in Dekalb County for three years in the late sixties. Gwinnett is so large that I shouldn’t generalize, but the parts you go through if you are so inclined to go to Atlanta are out of my worst nightmare.)
So here’s what I did, and you can suffer through the analysis or you can skip down to the bottom couple of paragraphs.
CCR gives the total number of stations - Oglethorpe, our very red county, is in the top four, even including Atlanta counties. But some stations are not reliable. CCR won’t count them as inactive unless the station operator says so, but in Oglethorpe, for instance, there are a couple of stations that gave a single report a year ago and then never again. CCR doesn’t drop any data, so those stations are still listed in the totals.
That’s why I have the column that I call “reliable.” It’s a judgement call, but if a station hasn’t reported in months, or if they make only intermittent reports when rain probably fell, if there are too many blanks when zeroes could have been reported, then I consider them unreliable. Oglethorpe County can usually count on 11 reliable reports, consistently, out of 19 stations.
The last three columns are all taken using July 2009 data, to date. I accepted CCR’s count of stations reporting Jul 1 to the present. The next to the last column uses that count as a rough estimate of the percent of reliable stations in the county. Mostly they hover aorund 60% of the total stations listed. Again, that’s mainly because CCR doesn’t delete stations. Oglethorpe maintains the highest percentage; Fulton County can only claim 40% of its total stations as being reliable. Our negative control, Floyd County, lost almost 50% of its stations to unreliability.
The last column is the percentage of stations reporting in July that I consider to be reliable. These generally run 80-90%, and Oglethorpe is right in there. But of the 61% reliable stations still in Dekalb County, 100% are reporting in July, and the same is true for Floyd County. Pretty impressive!
| County | Total | Reliable | Reporting in July | % “Reliable” (of “Total”) | % “Reliable” (of those "Reporting in July") |
|---|---|---|---|---|---|
| Oglethorpe | 19 | 11 | 13 | 68 | 85 |
| Clarke | 15 | 8 | 9 | 60 | 89 |
| Dekalb | 18 | 11 | 11 | 61 | 100 |
| Gwinnett | 23 | 12 | 14 | 61 | 86 |
| Fulton | 25 | 8 | 10 | 40 | 80 |
| Cobb | 18 | 8 | 10 | 56 | 80 |
| Chatham | 24 | 12 | 16 | 67 | 75 |
| Floyd | 11 | 6 | 6 | 55 | 100 |
What to make of all this? In populous counties in Georgia, at least, only around 60% of CCR observers stick with it reliably, over the course of a year. Of those, 80-90% reporting in the last month are in the reliable group, leaving 10-20% of the total stations that occasionally give a report but can’t be counted on. I guess that would leave 20-40% who totally dropped out after a short time.
I don’t know how well CoCoRaHS represents volunteer organizations in general, but it fits some of the parameters. People are initially attracted by some point of interest. 20-40% drop out in the first year. Of the ones remaining, 10-20% are unreliable. In the end, you can count on about half the initial interested sticking with you, at least through the first year.
