creating aeronet monthly data

4 posts / 0 new
Last post
Masaru Yoshioka
creating aeronet monthly data

Following your advice in my previous post, I now downloaded ALL_POINTS data.

So what do you think is a good strategy to compare monthly model outputs with AERONET AODs? I thought I may want to aggregate these point data for every month so I tried it like this way;

cis aggregate -v AOT_500:920801_171209_Izana.lev20 t=[1992-01-01T00:00:00,2017-12,P1M] -o AOT_500_Izana_monthly_920801_171209.nc

and this seems to have gone successfully, and the first couple of values are similar to the monthly data... well but actually the first data seems to be for July instead of June as I expected;

AOT_500 =
{_, _, _, _, _, _, _, _, _, _, _, _, _, _, _, _, _, _, _, _, _, _, _, _, _, # the last _ is for January 1994
_, _, _, _, _, _, _, _, _, _, _, _, _, _, _, _, _, _, _, _, _, _, _, _, # the last _ is for January 1996
_, _, _, _, _, _, _, _, _, _, _, _, _, _, _, _, _, 0.0117075107344633, # the last value is for July 1997
0.0417539232673267, _, _, _, _, _, _, _, _, _, _, _, _, _, _, _, _, _, _, # the first value is for August 1997

Values in the monthly mean data (AOT_500 of Daily Avg) are 0.011310, 0.042982 and N/A for June, July and August 1997, respectively. The very first data available in the point data is for 17:06:1997 as well.

I thought probably I only need 1995 and onward, so I did this;

cis aggregate -v AOT_500:920801_171209_Izana.lev20 t=[1995-01-01T00:00:00,2017-12,P1M] -o AOT_500_Izana_monthly_1995-2017.nc

Now I got this;

AOT_500 =
{_, _, _, _, _, _, _, _, _, _, _, _, _, _, _, _, _, _, _, _, _, _, _, _, _, # the last _ is for January 1997
_, _, _, _, 0.0131256116504854, 0.0300849257907543, 0.0290152742857143, #three values are for June, July and August 1997

So now the values have changed for some reason, and there are values for June, July and August 1997. However, neither the monthly nor the point data has value in August 1997 at all. Also, although the first value is very similar to the one in the monthly data, second data is a bit different. These must be a problem.

Do you see any problem in my calculation? Do you have any other advice on this? I'm doing these on JASMIN and data are stored the crescendo group work space. Thanks.

Best regards,
Masaru

Masaru Yoshioka
I checked the values of time

I checked the values of time along with AOT. In the first output,

AOT_500 =
{_, _, _, _, _, _, _, _, _, _, _, _, _, _, _, _, _, _, _, _, _, _, _, _, _, # 1-25
_, _, _, _, _, _, _, _, _, _, _, _, _, _, _, _, _, _, _, _, _, _, _, _, #26-49
_, _, _, _, _, _, _, _, _, _, _, _, _, _, _, _, _, 0.0117075107344633, #50-67
0.0417539232673267, _, _, _, _, _, _, _, _, _, _, _, _, _, _, _, _, _, _, #68-

time = 143190, 143220, 143250, 143280, 143310, 143340, 143370, 143400, # 1- 8
143430, 143460, 143490, 143520, 143550, 143580, 143610, 143640, 143670, # 9-17
143700, 143730, 143760, 143790, 143820, 143850, 143880, 143910, 143940, #18-26
143970, 144000, 144030, 144060, 144090, 144120, 144150, 144180, 144210, #27-35
144240, 144270, 144300, 144330, 144360, 144390, 144420, 144450, 144480, #36-44
144510, 144540, 144570, 144600, 144630, 144660, 144690, 144720, 144750, #45-53
144780, 144810, 144840, 144870, 144900, 144930, 144960, 144990, 145020, #54-62
145050, 145080, 145110, 145140, 145170, 145200, 145230, 145260, 145290, #63-71

so two values are at 145170 and 145200. On the other hand, in the second output,

AOT_500 =
{_, _, _, _, _, _, _, _, _, _, _, _, _, _, _, _, _, _, _, _, _, _, _, _, _, # 1-25
_, _, _, _, 0.0131256116504854, 0.0300849257907543, 0.0290152742857143, #26-32

time = 144286, 144316, 144346, 144376, 144406, 144436, 144466, 144496, # 1 - 8
144526, 144556, 144586, 144616, 144646, 144676, 144706, 144736, 144766, # 9 -17
144796, 144826, 144856, 144886, 144916, 144946, 144976, 145006, 145036, #18-26
145066, 145096, 145126, 145156, 145186, 145216, 145246, 145276, 145306, #27-35

So time for the 3 values are 145156, 145186, 145216.

OK, now I know what was the problem. P1M actually specifies 30 day period. So months would be offset by 5 or 6 days a year. So probably I should create a file for at least every 12 months, or I may want to create monthly files and concatenate them over months and maybe years.

Now I've got a new small question. How do I specify a period of 28 days for example in place of P1M? Can it be P28D? I tried this

cis aggregate -v AOT_500:920801_171209_Izana.lev20 t=[1997-06-01T00:00:00,1997-08,P30D] -o AOT_500_Izana_monthly_1997-0608.nc

and the results now look good.

Thanks,
Masaru

duncanwp
Yes, P28D will give you a

Yes, P28D will give you a period of 28 days. I'm glad that worked!

Masaru Yoshioka
Thanks a lot for your help!

Thanks a lot for your help!

Website designed & built by OCC