|
|
Post by Steelernation on Jan 5, 2019 19:56:36 GMT -5
Many US stations measure solar radiation while only 3 (that I know of measure sunshine). Is there any conversion between the two and if so, would it even be accurate?
Also, is there a way to find something like maximum possible solar radiation and create a percentage based off that?
The Oklahoma mesonet has data for percent sunshine that seems pretty accurate and aligns pretty well with the averages and what you would expect from a month. Not sure how that is calculated, there’s no data on sunshine minutes or hours and the other options for sunshine data are solar radiation...
It seems like it wouldn’t be that hard to do like actual solar radiation/max possible solar radiation but with all the data for solar radiation, the only percent sunshine I can find anywhere is the oklahoma mesonet.
Sunshine hours were recorded in Buffalo in 2015–the recorder was moved to buffalo state after the NWS stopped collecting but I haven’t found any data except for January-June 2015. I’d imagine it was collected for 6 months but I can’t find the data anywhere.
The lack of sunshine hours in the US is very frustrating so it would be nice to have a relatively accurate substitute.
|
|
|
|
Post by Crunch41 on Jan 6, 2019 1:19:36 GMT -5
I too would like to know this.
|
|
|
|
Post by jgtheone on Jan 6, 2019 3:37:31 GMT -5
Would love to know this too, as almost every BoM station records solar radiation, but only major ones and some anomalous stations record sunshine (i.e. Eildon)
|
|
|
|
Post by Donar on Jan 6, 2019 6:28:03 GMT -5
|
|
|
|
Post by bizzy on Jan 6, 2019 10:58:17 GMT -5
Sunshine hours are simply the amount of time the solar radiation is at/above 120w/m^2, so it shouldn’t be too difficult to convert, the most difficult part would probably be retrieving the data.
Also, I’m not sure how solar radiation is recorded in the US (I didn’t even know it was recorded tbh), nor the frequency in which it’s recorded, the standard hourly updates issued by the NWS are mostly sufficient for weather conditions, but I don’t know how that’d work for sunshine hours, especially for areas that are generally partly cloudy (most of the US). 59 minutes of sunshine + 1 minute of heavy clouds at the wrong time could result in a loss of an entire hour of sunshine.
|
|
|
|
Post by Steelernation on Jan 6, 2019 18:07:18 GMT -5
Also, I’m not sure how solar radiation is recorded in the US (I didn’t even know it was recorded tbh), nor the frequency in which it’s recorded, the standard hourly updates issued by the NWS are mostly sufficient for weather conditions, but I don’t know how that’d work for sunshine hours, especially for areas that are generally partly cloudy (most of the US). Many state mesonets measure it as like mean daily solar radiation so that should be more frequent than every year. I don’t think the NWS measures it, at least not that I’ve found. It would seem simple to find all the hours over 120 w/m2 but then you’d have to go through all the data for each day and add it up which would be very hard and painstaking. Might be some excel program that could do that but I’m not very good with software like that. |
|
|
|
Post by nei on Jan 6, 2019 20:27:17 GMT -5
It would seem simple to find all the hours over 120 w/m2 but then you’d have to go through all the data for each day and add it up which would be very hard and painstaking. Might be some excel program that could do that but I’m not very good with software like that. eh, that's a less complicated task that my precipitation plots. Doesn't sound that hard, but where's the data? I remember solar data available on a few stations available from the NCDC, but the government shutdown means the NCDC because it's not an essential service according to the powers that be. |
|
|
|
Post by nei on Jan 6, 2019 20:52:24 GMT -5
|
|
|
|
Post by Steelernation on Jan 6, 2019 21:07:46 GMT -5
eh, that's a less complicated task that my precipitation plots. Doesn't sound that hard, but where's the data? I remember solar data available on a few stations available from the NCDC, but the government shutdown means the NCDC because it's not an essential service according to the powers that be. For New York, the ny mesonet measures solar radiation. You can request 5 minute solar radiation data here for free. Here there’s a graph of it for a day but it doesn’t appear to go back anymore. Other state climatologists/state mesonets have solar radiation data. A lot of it is mean daily which might work, there might be 1 minute or 5 minute data on some of them but I didn’t look. I assume you’d need like 1 minute or 5 minute data for that? Maybe a daily graph would work if you could find graphs for past days. |
|
|
|
Post by nei on Jan 6, 2019 23:21:29 GMT -5
eh, that's a less complicated task that my precipitation plots. Doesn't sound that hard, but where's the data? I remember solar data available on a few stations available from the NCDC, but the government shutdown means the NCDC because it's not an essential service according to the powers that be. For New York, the ny mesonet measures solar radiation. You can request 5 minute solar radiation data here for free. Here there’s a graph of it for a day but it doesn’t appear to go back anymore. Other state climatologists/state mesonets have solar radiation data. A lot of it is mean daily which might work, there might be 1 minute or 5 minute data on some of them but I didn’t look. I assume you’d need like 1 minute or 5 minute data for that? Maybe a daily graph would work if you could find graphs for past days. I'm going to request 5 minute data, any particular places you'd like? |
|
|
|
Post by nei on Jan 7, 2019 10:39:42 GMT -5
For New York, the ny mesonet measures solar radiation. You can request 5 minute solar radiation data here for free. Here there’s a graph of it for a day but it doesn’t appear to go back anymore. Other state climatologists/state mesonets have solar radiation data. A lot of it is mean daily which might work, there might be 1 minute or 5 minute data on some of them but I didn’t look. I assume you’d need like 1 minute or 5 minute data for that? Maybe a daily graph would work if you could find graphs for past days. I'm going to request 5 minute data, any particular places you'd like? Downloaded 4 NY stations, it's one csv file per day, which is kinda annoying |
|
|
|
Post by Steelernation on Jan 7, 2019 15:14:04 GMT -5
Downloaded 4 NY station, it's one csv file per day, which is kinda annoying That does sound annoying, probably makes it very time consuming. If you didn’t already could you do Rochester? I don’t think there’s a Rochester station, Rush is the closest one. |
|
|
|
Post by nei on Jan 7, 2019 15:36:44 GMT -5
Downloaded 4 NY station, it's one csv file per day, which is kinda annoying That does sound annoying, probably makes it very time consuming. If you didn’t already could you do Rochester? I don’t think there’s a Rochester station, Rush is the closest one. I downloaded Brockport, thinking that was the closest to Rochester. What do you want? Monthly sunshine hours? Percent? Figured out how to get it to loop over all files, not too bad, just takes the computer a while to load each file. |
|
|
|
Post by Steelernation on Jan 7, 2019 16:05:46 GMT -5
I downloaded Brockport, thinking that was the closest to Rochester. What do you want? Monthly sunshine hours? Percent? Figured out how to get it to loop over all files, not too bad, just takes the computer a while to load each file. Thanks, Brockport is just as close. Either one works for me, pretty easy for me to convert from hours to percent or visa versa. |
|
|
|
Post by Babu on Jan 11, 2019 16:30:13 GMT -5
Is the solar radiation measured at a perpendicular angle towards the sun? Otherwise, the answer is simply no.
And even if it is, you'd have to have very frequent data. If you have hourly averages, how are you going to distinguish between 50 minutes of thin overcast, or 10 minutes of thick overcast?
|
|
|
|
Post by Steelernation on Jan 11, 2019 17:04:42 GMT -5
Is the solar radiation measured at a perpendicular angle towards the sun? Otherwise, the answer is simply no. And even if it is, you'd have to have very frequent data. If you have hourly averages, how are you going to distinguish between 50 minutes of thin overcast, or 10 minutes of thick overcast? Not sure how is measured. The data is 5 minute data so that should be pretty accurate. |
|
|
|
Post by Steelernation on Jan 12, 2019 17:08:07 GMT -5
nei have you figured out what your mistake was? Or is the data just inaccurate and not usable?
|
|
|
|
Post by nei on Jan 12, 2019 18:03:08 GMT -5
nei have you figured out what your mistake was? Or is the data just inaccurate and not usable? Never got around to checking it. For those not following the shoutbox conversation then, I counted the number of 5-minute intervals where solar radiation was above 120 W/m^2. Got 2600-2700 hours for a station near Rochester for 2017 & 2018. Either I made a mistake, data is just inaccurate or maybe you can't solar radiation to sunshine hours. Best would be to download Blue Hill Observatory from the same source and see how it compares to the correct values. But your link is only for NY. It's got to be a subtle error to be only off by like 20%; most mistakes I would think would do something larger. |
|
|
|
Post by boombo on Jan 26, 2019 18:27:14 GMT -5
nei have you figured out what your mistake was? Or is the data just inaccurate and not usable? Never got around to checking it. For those not following the shoutbox conversation then, I counted the number of 5-minute intervals where solar radiation was above 120 W/m^2. Got 2600-2700 hours for a station near Rochester for 2017 & 2018. Either I made a mistake, data is just inaccurate or maybe you can't solar radiation to sunshine hours. Best would be to download Blue Hill Observatory from the same source and see how it compares to the correct values. But your link is only for NY. It's got to be a subtle error to be only off by like 20%; most mistakes I would think would do something larger. Maybe I'm missing something here but surely solar radiation depends on what time of day/time of year the sun's out while raw sunshine hours doesn't take that into account, so they can't be convertable? Where I live we get a lot of days in late spring/early summer with the dreaded sunny morning/cloudy afternoon/sunny evening combination, so we're still getting the amount of sun hours we're supposed to, just not when it's at its strongest. |
|
|
|
Post by nei on Jan 27, 2019 22:09:03 GMT -5
Never got around to checking it. For those not following the shoutbox conversation then, I counted the number of 5-minute intervals where solar radiation was above 120 W/m^2. Got 2600-2700 hours for a station near Rochester for 2017 & 2018. Either I made a mistake, data is just inaccurate or maybe you can't solar radiation to sunshine hours. Best would be to download Blue Hill Observatory from the same source and see how it compares to the correct values. But your link is only for NY. It's got to be a subtle error to be only off by like 20%; most mistakes I would think would do something larger. Maybe I'm missing something here but surely solar radiation depends on what time of day/time of year the sun's out while raw sunshine hours doesn't take that into account, so they can't be convertable? Where I live we get a lot of days in late spring/early summer with the dreaded sunny morning/cloudy afternoon/sunny evening combination, so we're still getting the amount of sun hours we're supposed to, just not when it's at its strongest. from what I've heard from other forum members the C-S recorders [usual ones outside the US] count sunshine as anytime the solar radiation is over 120 Watts per square meter. The data from Steelernation 's link was in 5 minute intervals, so counted the number of times the solar radiation ≥ 120 and multiplied by 5/60 to get sunshine hours. Maybe it's more complicated than that, but I can't bothered to do work on it. Cloudiest in afternoon does sound frustrating; the US west coast mostly has the opposite of that in summer, where I lived in Massachusetts is prone to valley fog so perhaps a little cloudier in morning; would be interesting to check |
|
