def site_iter_process(valid_refs,c):
#for r in range(len(valid_refs)):
ref = valid_refs[c]
print ref
#get site instrument for species
met_i = met_refs.index(ref)
print len(met_refs)
print len(met_species)
site_species = list(met_species[met_i])
site_instruments = list(met_instruments[met_i])
print site_species
print site_instruments
mm = site_instruments[site_species.index(species)]
site_resolutions = []
data_valid = True
s_files = insensitive_glob('/work/home/db876/observations/surface/%s/EANET/*%s.csv'%(species,ref))
site_files = []
for y in year_array:
for f in s_files:
if str(y)[-2:] in f:
site_files.append(f)
site_files = modules.natsorted(site_files)
if site_files == []:
print 'No files for ref.\n'
years = []
months = []
days = []
hours = []
vals = []
last_year_index = len(site_files)
for y in year_array:
print 'Processing Year %s'%y
got_year = False
for file in site_files:
last_file_split = file.split('/')[-1]
if str(y)[2:] in last_file_split:
got_year = True
break
if got_year == False:
#fill in data for missing year
timedelta_diff = datetime.date(y+1, 1, 1) - datetime.date(y, 1, 1)
ndays_missing = timedelta_diff.days
print 'ndays missing = ', ndays_missing
vals=np.append(vals,[-99999]*(ndays_missing*24))
continue
print file
valid = True
with open(file, 'rb') as f:
reader = csv.reader(f,delimiter=',')
counter = 0
#get resolution
for row in reader:
if counter == 0:
all_units = row
elif counter == 1:
file_res = 'H'
try:
hour_index = row.index('Hour')
except:
file_res = 'D'
try:
day_index = row.index('Day')
except:
file_res = 'M'
month_index = row.index('Month')
year_index = row.index('Year')
try:
spec_index = row.index(species.upper())
units = all_units[spec_index]
except:
valid = False
break
#make sure each year units are ppb
if units != 'ppb':
print 'Units not ppb!'
1+'a'
if counter == 2:
if file_res == 'H':
yyyy = row[year_index]
mm = row[month_index]
dd = row[day_index]
hh = row[hour_index]
elif file_res == 'D':
yyyy = row[year_index]
mm = row[month_index]
dd = row[day_index]
hh = 1
elif file_res == 'M':
yyyy = row[year_index]
mm = row[month_index]
dd = 1
hh = 1
start_datetime = datetime.datetime(int(yyyy),int(mm),int(dd),int(hh))
if counter == 3:
if file_res == 'H':
yyyy = row[year_index]
mm = row[month_index]
dd = row[day_index]
hh = row[hour_index]
elif file_res == 'D':
yyyy = row[year_index]
mm = row[month_index]
dd = row[day_index]
hh = 1
elif file_res == 'M':
yyyy = row[year_index]
mm = row[month_index]
dd = 1
hh = 1
present_datetime = datetime.datetime(int(yyyy),int(mm),int(dd),int(hh))
time_delt = present_datetime-start_datetime
hour_delt = datetime.timedelta(hours=1)
day_delt = datetime.timedelta(hours=24)
week_delt = datetime.timedelta(hours=24*7)
month_delt = datetime.timedelta(hours=24*28)
print time_delt
if (time_delt < day_delt):
print 'Hourly Data'
file_res = 'H'
site_resolutions.append(file_res)
elif (time_delt > hour_delt) & (time_delt < week_delt):
print 'Daily Data'
file_res = 'D'
site_resolutions.append(file_res)
elif (time_delt > week_delt):
print 'Monthly Data'
file_res = 'M'
site_resolutions.append(file_res)
#break
#limit files by timeres return if not suitable for output res
if output_res == 'H':
if (file_res == 'D') or (file_res == 'M'):
print 'File resolution has to be Minimum Hourly. Skipping'
data_valid = False
return c,vals,data_valid,-999,-999,-999,'na','na','na','na','na'
elif (output_res == 'D'):
if (file_res == 'M'):
print 'File resolution has to be Minimum Daily. Skipping'
data_valid = False
return c,vals,data_valid,-999,-999,-999,'na','na','na','na','na'
counter+=1
#READ IN DATA
if valid == True:
with open(file, 'rb') as f:
reader = csv.reader(f,delimiter=',')
counter = 0
for row in reader:
if counter >= 2:
yyyy = row[year_index]
mm = row[month_index]
if file_res == 'H':
try:
vals=np.append(vals,np.float64(row[spec_index]))
except:
vals=np.append(vals,-99999)
elif file_res == 'D':
try:
vals=np.append(vals,[np.float64(row[spec_index])]*24)
except:
vals=np.append(vals,[-99999]*24)
elif file_res == 'M':
month_days = monthrange(int(yyyy), int(mm))[1]
try:
vals=np.append(vals,[np.float64(row[spec_index])]*(month_days*24))
except:
vals=np.append(vals,[-99999]*(month_days*24))
counter+=1
else:
print 'Species is not in file header. Skipping Year'
timedelta_diff = datetime.date(y+1, 1, 1) - datetime.date(y, 1, 1)
ndays_missing = timedelta_diff.days
print 'ndays missing = ', ndays_missing
vals=np.append(vals,[-99999]*(ndays_missing*24))
valid_refs_rev.append(ref)
i_ref = met_refs.index(ref)
tz = np.float64(met_tz[i_ref])
lat = np.float64(met_lats[i_ref])
lon = np.float64(met_lons[i_ref])
alt = np.float64(met_alts[i_ref])
raw_class_name = met_class[i_ref]
anthrome_class_name = class_name[i_ref]
#check tz is whole number else skip site
if (tz % 1) != 0:
print 'Timezone is not even. Skipping'
data_valid = False
tz = int(tz)
#correct time to UTC
if tz < 0:
#get rid of values at start and append -99999's at end
cut = vals[:tz]
for num in range(np.abs(tz)):
cut = np.insert(cut,0, -99999)
vals = cut
elif tz > 0:
#put -99999's at start and get rid of values at end
cut = vals[tz:]
for num in range(tz):
cut = np.append(cut, -99999)
vals = cut
#do data quality checks
full_data,data_valid = modules.quality_check(vals,data_valid,data_resolution,np.float64(alt),grid_dates,start_year,end_year)
#if all site resolutions are same continue, make program exit
all_same = all(x == site_resolutions[0] for x in site_resolutions)
if all_same == True:
pass
else:
print 'Not all files for site have same resolution. Skipping.'
data_valid = False
return c,full_data,data_valid,-999,-999,-999,'na','na','na','na','na'
#set sampling as average
st = 'average'
return c,full_data,data_valid,lat,lon,alt,raw_class_name,anthrome_class_name,mm,st,file_res
def site_iter_process(valid_refs, c):
# for each valid location process
# limit obs data due for each site in valid_obs_site_names
# for c in range(len(valid_refs)):
all_lat = []
all_lon = []
all_alt = []
all_st = []
all_mm = []
site_ref = valid_refs[c]
file_valid = True
data_valid = True
print site_ref
file_res = data_resolutions[c]
print file_res
# read files for each valid site
s_files = sorted(
glob.glob("/work/home/db876/observations/surface/%s/GAW/%s**.%s**.dat" % (species, site_ref.lower(), file_res))
)
print s_files
if file_res == "hr":
site_files = sorted(s_files, key=lambda x: x.split(".hr")[1])
else:
site_files = sorted(s_files)
delete_inds = []
if file_res == "hr":
# limit site files before and after year limit
for i in range(len(site_files)):
f = site_files[i]
year = f.split(".hr")[1][:4]
if int(year) < int(start_year):
delete_inds.append(i)
if int(year) > int(end_year):
delete_inds.append(i)
site_files = np.delete(site_files, delete_inds)
print site_files
site_file_len = len(site_files)
s_count = 0
start_ind = 0
end_ind = 0
for f in site_files:
print f
read = np.loadtxt(f, dtype="S10,S5,f8", comments="C", usecols=(0, 1, 4), unpack=True)
read = np.array(read)
dates = read[0, :]
times = read[1, :]
conc = read[2, :]
conc = np.array(conc)
conc = conc.astype(float)
# change all vals < 0 to np.NaN
inv_test = conc < 0
conc[inv_test] = np.NaN
start_ind = end_ind
end_ind += len(conc)
s_count += 1
units = []
mycsv = csv.reader(open(f))
row_count = 0
for row in mycsv:
if row_count == 11:
val = " ".join(row)
lat = val.replace(" ", "")
lat = lat[12:]
lat = float(lat)
all_lat.append(lat)
# get lon
if row_count == 12:
val = " ".join(row)
lon = val.replace(" ", "")
lon = lon[13:]
lon = float(lon)
all_lon.append(lon)
# get altitude
if row_count == 13:
val = " ".join(row)
alt = val.replace(" ", "")
alt = alt[12:]
alt = float(alt)
all_alt.append(alt)
# get units
if row_count == 20:
val = " ".join(row)
unit = val.replace(" ", "")
unit = unit[19:]
# get measurement method
if row_count == 21:
val = " ".join(row)
mm = val.replace(" ", "")
mm = mm[21:]
all_mm.append(mm)
# get sampling type
if row_count == 22:
val = " ".join(row)
st = val.replace(" ", "")
st = st[16:]
all_st.append(st)
if row_count == 23:
val = " ".join(row)
tz = val.replace(" ", "")
tz = tz[12:]
row_count += 1
# test if units are in ppb for each file - if not convert
if (unit != "ppb") & (unit != "ppbv"):
if (unit == "ug/m3") or (unit == "ugN/m3"):
print "converting units, temp = 20degC"
# calculate conversion factor from mg/m3 assuming 20 degC and 1 atm - default for GAW site O3 instruments
# R/MW*(TEMP0C(K)*TEMP(degC)/P(hPa)/10
conv_fact = 8.3144 / mol_mass * (273.15 + 20) / (1013.25 / 10)
conc = conv_fact * conc
elif (unit == "ug/m3-20C") or (unit == "ugN/m3-20C"):
print "converting units, temp = 20degC"
# calculate conversion factor from mg/m3 assuming 20 degC and 1 atm - default for GAW site O3 instruments
# R/MW*(TEMP0C(K)*TEMP(degC)/P(hPa)/10
conv_fact = 8.3144 / mol_mass * (273.15 + 20) / (1013.25 / 10)
conc = conv_fact * conc
elif (unit == "ug/m3-25C") or (unit == "ugN/m3-25C") or (unit == "ug/m3at25C"):
print "converting units, temp = 25degC"
# calculate conversion factor from mg/m3 assuming 25 degC and 1 atm
# R/MW*(TEMP0C(K)*TEMP(degC)/P(hPa)/10
conv_fact = 8.3144 / mol_mass * (273.15 + 25) / (1013.25 / 10)
conc = conv_fact * conc
elif (unit == "mg/m3-20C") or (unit == "mgN/m3-20C"):
print "converting units, temp = 25degC"
# calculate conversion factor from mg/m3 assuming 25 degC and 1 atm
# R/MW*(TEMP0C(K)*TEMP(degC)/P(hPa)/10
conv_fact = 8.3144 / mol_mass * (273.15 + 20) / (1013.25 / 10)
conc = (conv_fact * conc) * 1e3
elif (unit == "mg/m3-25C") or (unit == "mgN/m3-25C"):
print "converting units, temp = 25degC"
# calculate conversion factor from mg/m3 assuming 25 degC and 1 atm
# R/MW*(TEMP0C(K)*TEMP(degC)/P(hPa)/10
conv_fact = 8.3144 / mol_mass * (273.15 + 25) / (1013.25 / 10)
conc = (conv_fact * conc) * 1e3
elif (unit == "ppm") or (unit == "ppmv"):
conc = conc * 1.0e3
elif (unit == "ppt") or (unit == "pptv"):
conc = conc / 1.0e3
else:
print "Unknown Unit"
print unit
1 + "a"
break
if tz != "UTC":
if tz == "":
if site_ref.lower() in ["plm"]:
tz = -5
if site_ref.lower() in ["kos", "edm", "vdl", "nwr"]:
tz = 0
if site_ref.lower() in [
"jfj",
"kps",
"rig",
"pay",
"glh",
"cmn",
"zep",
"dig",
"hhe",
"ktb",
"stp",
"ivn",
"jcz",
"kam",
"lzp",
"snz",
"zbl",
"kmw",
"don",
"mhn",
"nia",
"roq",
"spm",
]:
tz = 1
if site_ref.lower() in ["rcv", "aht", "oul", "uto", "vir", "fdt", "sem", "stn"]:
tz = 2
if site_ref.lower() in ["dak"]:
tz = 3
if site_ref.lower() in ["shp"]:
tz = 4
if site_ref.lower() in ["isk"]:
tz = 5
if site_ref.lower() in ["hkg"]:
tz = 8
if site_ref.lower() in ["cgo"]:
tz = 10
else:
tz = tz.replace("LocaltimeUTC", "")
tz = tz.replace("OtherUTC", "")
tz = tz.replace("Localtime", "")
tz = tz.replace(":", ".")
try:
before, sep, after = tz.rpartiton(".")
after = int(after)
conv = (100.0 / 60) * after
tz = before + sep + str(conv)
except:
1 + 1
tz = float(tz)
else:
tz = 0
# check tz is whole number else skip site
if (tz % 1) != 0:
print "File Invalid, timezone is not a whole number."
conc[:] = -99999
# process dates from date, time to days since start year
dates = [s.replace("-", "") for s in dates]
times = [s.replace(":", "") for s in times]
if file_res == "hr":
# some times go from 0100 to 2400, assume this is when sites report ave for hour previous. Thus all times should have hour minused
for i in range(len(times)):
if times[i] == "2400":
current_date = dates[i]
test = np.array(dates) == current_date
indices = [i for i, x in enumerate(test) if x]
for x in indices:
current_time = times[x]
if current_time == "2400":
current_time = "0000"
date_datetime = datetime.datetime(
int(current_date[0:4]),
int(current_date[4:6]),
int(current_date[6:]),
int(current_time[:2]),
int(current_time[2:]),
)
date_datetime = date_datetime - datetime.timedelta(hours=1)
times[x] = date_datetime.strftime("%H%M")
# adjust dates and times if tz is not equal to 0
if tz != 0:
for i in range(len(dates)):
# create datetime
dt = datetime.datetime(
int(dates[i][:4]), int(dates[i][4:6]), int(dates[i][6:]), int(times[i][:2]), int(times[i][2:])
)
if tz > 0:
# print 'Old dt', dt
dt = dt - datetime.timedelta(hours=int(tz))
# print 'New dt', dt
elif tz < 0:
# print 'Old dt', dt
dt = dt + datetime.timedelta(hours=np.abs(int(tz)))
# print 'New dt', dt
dates[i] = dt.strftime("%Y%m%d")
times[i] = dt.strftime("%H%M")
data = [dates, times, conc]
try:
big_list = np.hstack((big_list, data))
except:
big_list = np.array(data)
if s_count == site_file_len:
# make sure big list exists
try:
big_list
except:
data_valid = False
if data_valid == True:
# get dates and times
date_con = big_list[0, :]
time_con = big_list[1, :]
# get vals
vals = np.array(big_list[2, :]).astype(float)
# delete big list
del big_list
# if dates outside what asked for exclude
first_date_val = int("%s0101" % (start_year))
last_date_val = int("%s1231" % (end_year))
test_valid = (np.array(date_con).astype(int) >= first_date_val) & (
np.array(date_con).astype(int) <= last_date_val
)
date_con = date_con[test_valid]
time_con = time_con[test_valid]
vals = vals[test_valid]
# Check if any times are duplicate, if so delete all but first
del_list = []
for d in range(len(date_con) - 1):
if (date_con[d] == date_con[d + 1]) & (time_con[d] == time_con[d + 1]):
del_list.append(d + 1)
if len(del_list) > 0:
print "Deleting duplicate timepoints"
print date_con[del_list], time_con[del_list]
date_con = np.delete(date_con, del_list)
time_con = np.delete(time_con, del_list)
vals = np.delete(vals, del_list)
# if file resolution is daily or monthly then replicate times after point, to fill hourly data array.
count = 0
if file_res == "da":
file_hours = len(date_con)
for i in range(file_hours):
current_hh = int(time_con[count][:2])
current_mm = int(time_con[count][2:])
s = datetime.datetime(year=start_year, month=1, day=1, hour=current_hh, minute=current_mm)
e = datetime.datetime(year=start_year, month=1, day=2, hour=current_hh, minute=current_mm)
day_hours = [d.strftime("%H%M") for d in pd.date_range(s, e, freq="H")][1:-1]
date_con = np.insert(date_con, count + 1, [date_con[count]] * 23)
time_con = np.insert(time_con, count + 1, day_hours)
vals = np.insert(vals, count + 1, [vals[count]] * 23)
count += 24
if file_res == "mo":
file_hours = len(date_con)
for i in range(file_hours):
current_year = int(date_con[count][:4])
current_month = int(date_con[count][4:6])
next_month = current_month + 1
if next_month > 12:
next_month = 1
next_year = current_year + 1
else:
next_year = current_year
s = datetime.datetime(year=current_year, month=current_month, day=1, hour=1, minute=0)
e = datetime.datetime(year=next_year, month=next_month, day=1, hour=0, minute=0)
day_date = [d.strftime("%Y%m%d") for d in pd.date_range(s, e, freq="H")][:-1]
day_hour = [d.strftime("%H%M") for d in pd.date_range(s, e, freq="H")][:-1]
date_con = np.insert(date_con, count + 1, day_date)
time_con = np.insert(time_con, count + 1, day_hour)
vals = np.insert(vals, count + 1, [vals[count]] * len(day_date))
count += len(day_date) + 1
date_con = np.array(date_con).astype(int)
time_con = np.array(time_con).astype(int)
# create max possible o3 grid
o3_data = np.empty(n_hours)
o3_data[:] = -99999
# delete dates,times and var outside date range
val_test = (date_con >= int(output_res_dates_strings[0])) & (
date_con <= int(output_res_dates_strings[-1])
)
date_con = date_con[val_test]
time_con = time_con[val_test]
vals = vals[val_test]
print date_con
# find matching times between actual times and grid of times, return big array of indices of matched indices in grid
converted_time = modules.date_process(date_con, time_con, start_year)
converted_time = np.round(converted_time, decimals=5)
syn_grid_time = np.arange(0, n_days, 1.0 / 24)
syn_grid_time = np.round(syn_grid_time, decimals=5)
# find matching times between actual times and grid of times, return big array of indices of matched indices in grid
indices = np.searchsorted(syn_grid_time, converted_time, side="left")
o3_data[indices] = vals
# convert all Nans back to -99999
test = np.isnan(o3_data)
o3_data[test] = -99999
# get mode of metadata
lat = np.float64(stats.mode(all_lat)[0][0])
lon = np.float64(stats.mode(all_lon)[0][0])
alt = np.float64(stats.mode(all_alt)[0][0])
st = stats.mode(all_st)[0][0]
mm = stats.mode(all_mm)[0][0]
# check site is not urban using anthrome map from 2000
anthfile = "/work/home/db876/plotting_tools/core_tools/anthro2_a2000.nc"
anthload = Dataset(anthfile)
class_valid, anthrome_class_name = modules.anthrome_classify(anthload, [lat], [lon])
if class_valid == "invalid":
data_valid = False
print "Site Invalid, site classed as urban by anthrome map."
# get measurement type and sampling type (take mode from collected list)
if (st == "continuous") or (
st == "continuous(carbondioxide),remotespectroscopicmethod(methaneandsurfaceozone)"
):
st = "average"
elif st == "flask":
st = "flask"
elif st == "filter":
st = "filter"
else:
print st
1 + "a"
if mm == "Lightabsorptionanalysis(UV)":
mm = "ultraviolet photometry"
elif mm == "CavityRingdownSpectroscopy":
mm = "cavity ringdown spectroscopy"
elif mm == "NDIR":
site_mm = "non-dispersive infrared spectroscopy"
elif mm == "GasChromatography(FID)":
site_mm = "gas chromatography flame ionisation detection"
elif mm == "Gas Chromatography (RGD)":
site_mm = "gas chromatography reduction gas detection"
elif mm == "Chemiluminescence":
mm = "chemiluminescence"
elif (mm == "Spectrophotometry") or (
mm == "spectrophotometry,naphthyl-ethylenediaminedihydrochloridemethod"
):
mm = "spectrophotometry"
elif mm == "":
if species == "O3":
mm = "ultraviolet photometry"
if species == "CO":
mm = "non-dispersive infrared spectroscopy"
if species == "NO2":
mm = "chemiluminescence"
if species == "NO":
mm = "chemiluminescence"
if species == "ISOP":
mm = "gas chromatography flame ionisation detection"
# do data quality checks
full_data, data_valid = modules.quality_check(
o3_data, data_valid, data_resolution, alt, grid_dates, start_year, end_year
)
# convert file res to standard format
if file_res == "hr":
file_res = "H"
elif file_res == "da":
file_res = "D"
elif file_res == "mo":
file_res = "M"
# no raw class so set as na
raw_class_name = "na"
return c, full_data, data_valid, lat, lon, alt, raw_class_name, anthrome_class_name, mm, st, file_res
def site_iter_process(valid_refs,c):
#for site_ref in valid_refs:
data_valid = True
site_ref = valid_refs[c]
print 'ref = ',site_ref
site_test = all_refs == site_ref
site_yyyymmdd = yyyymmdd[site_test]
site_hhmm = hhmm[site_test]
site_vals = vals[site_test]
#convert blank invalids to -99999
test_inv = site_vals == ''
site_vals[test_inv] = -99999
site_vals = np.float64(site_vals)
#convert number invalids to -99999
test_inv = site_vals < 0
site_vals[test_inv] = -99999
#put vals into full grid
date_con = np.array(site_yyyymmdd).astype(int)
time_con = np.array(site_hhmm).astype(int)
#create max possible o3 grid
full_data = np.empty(n_hours)
full_data[:] = -99999
#find matching times between actual times and grid of times, return big array of indices of matched indices in grid
converted_time = modules.date_process(date_con,time_con,start_year)
converted_time = np.round(converted_time,decimals=5)
syn_grid_time = np.arange(0,n_days,1./24)
syn_grid_time = np.round(syn_grid_time,decimals=5)
#find matching times between actual times and grid of times, return big array of indices of matched indices in grid
indices = np.searchsorted(syn_grid_time, converted_time, side='left')
site_vals = np.array(site_vals)
full_data[indices] = site_vals
meta_index = meta_refs.index(site_ref)
tz = meta_tz[meta_index]
lat = np.float64(meta_lats[meta_index])
lon = np.float64(meta_lons[meta_index])
alt = np.float64(meta_alts[meta_index])
raw_class_name = meta_class[meta_index]
anthrome_class_name = class_name[meta_index]
#correct timezone to UTC
if tz < 0:
#get rid of values at start and append -99999's at end
cut = full_data[:tz]
for num in range(np.abs(tz)):
cut = np.insert(cut,0, -99999)
full_data = cut
elif tz > 0:
#put -99999's at start and get rid of values at end
cut = full_data[tz:]
for num in range(tz):
cut = np.append(cut, -99999)
full_data = cut
#do data quality checks
full_data,data_valid = modules.quality_check(full_data,data_valid,data_resolution,alt,grid_dates,start_year,end_year)
#set mm
if species == 'O3':
mm = 'ultraviolet photometry'
elif (species == 'NO') or (species == 'NO2') or (species == 'CO'):
mm = 'non-dispersive infrared absorption'
#set sampling as average
st = 'average'
#set site file resolution
file_res = 'H'
return c,full_data,data_valid,lat,lon,alt,raw_class_name,anthrome_class_name,mm,st,file_res
def site_iter_process(valid_refs,c):
#read files site at a time
#for ref_i in range(len(valid_refs)):
site_ref = valid_refs[c]
all_latitudes = []
all_longitudes = []
all_altitudes = []
all_mm = []
print 'Current Ref is = ', site_ref
#find if sites have full valid range from start year and finishing in end year
s_files = glob.glob('/work/home/db876/observations/surface/%s/EMEP/%s*'%(species,site_ref))
year_files = [file.replace("/work/home/db876/observations/surface/%s/EMEP/"%(species), "") for file in s_files]
cut_year_files = [file[8:12] for file in year_files]
site_files = []
for y in year_array:
for i in range(len(s_files)):
if str(y) in cut_year_files[i]:
site_files.append(s_files[i])
site_files = modules.natsorted(site_files)
year_files = modules.natsorted(year_files)
file_startdate = []
file_height = []
instr_names = []
file_lasttime = []
data_valid = True
yyyymmdd = []
hhmm = []
vals = []
flags = []
#create max possible o3 grid
full_data = np.empty(n_hours)
full_data[:] = -99999
if site_files == []:
print 'No valid files for site\n'
return
for y in year_array:
print 'Processing Year %s'%y
got_year = False
for file in site_files:
last_file_split = file.split('/')[-1]
if str(y) in last_file_split[8:12]:
got_year = True
break
if got_year == False:
#fill in data for missing year
timedelta_diff = datetime.date(y+1, 1, 1) - datetime.date(y, 1, 1)
ndays_missing = timedelta_diff.days
print 'ndays missing = ', ndays_missing
continue
if data_valid == True:
data_start = 9999999
count = 0
start_read = False
with open(file, 'rb') as f:
read_count = 0
reader = csv.reader(f,delimiter=' ')
print file
for row in reader:
try:
row = filter(lambda a: a != '', row)
except:
pass
try:
row = filter(lambda a: a != ',', row)
except:
pass
#get start date of file
if row[0] == 'Startdate:':
data = row[1]
s_yyyy = data[:4]
s_mm = data[4:6]
s_dd = data[6:8]
s_hh = data[8:10]
s_min = data[10:12]
start_datetime = datetime.datetime(int(s_yyyy),1,1,0,0)
#get unit
if row[0] == 'Unit:':
try:
unit_part1 = row[1]
unit_part2 = row[2]
unit = unit_part1+'_'+unit_part2
except:
unit = row[1]
#get resolution
if row[0] == 'Resolution':
if row[1] == 'code:':
file_res = row[2]
print 'Resolution = %s'%file_res
if (output_res == 'H'):
if (file_res == '1d') or (file_res == '1mo'):
print 'File resolution has to be Minimum Hourly. Skipping'
data_valid = False
return c,full_data,data_valid,-999,-999,-999,'na','na','na','na','na'
elif (output_res == 'D'):
if (file_res == '1mo'):
print 'File resolution has to be Minimum Daily. Skipping'
data_valid = False
return c,full_data,data_valid,-999,-999,-999,'na','na','na','na','na'
#get latitude
if row[0] == 'Station':
if row[1] == 'latitude:':
latitude = row[2]
all_latitudes.append(latitude)
#get longitude
if row[0] == 'Station':
if row[1] == 'longitude:':
longitude = row[2]
all_longitudes.append(longitude)
#get altitude
if row[0] == 'Station':
if row[1] == 'altitude:':
altitude = row[2][:-1]
all_altitudes.append(altitude)
#get period
if row[0] == 'Period':
period_code = row[2]
#get stats method
if row[0] == 'Statistics:':
try:
st = row[1] + row[2]
if st != 'arithmeticmean':
print 'Not Arithmetic Mean!'
print row[1]
print 1+'a'
except:
print 'Not Arithmetic Mean!'
print row[1]
print 1+'a'
#get instrument method
if row[0] == 'Instrument':
if row[1] == 'type:':
mm_list = row[2:]
if len(mm_list) > 1:
site_mm = ''
for x in range(len(mm_list)):
site_mm = site_mm+mm_list[x]+' '
site_mm = site_mm.strip()
else:
site_mm = mm_list[0]
all_mm.append(site_mm)
#get data
if start_read == True:
#calc dates, times, and take o3 vals
time_since_start = np.float64(row[0])
days_since_start = math.trunc(time_since_start)
remainder = time_since_start - days_since_start
unrounded_hour = remainder*24
hour = np.round(unrounded_hour)
time_delta = datetime.timedelta(days = days_since_start,hours = hour)
calc_datetime = start_datetime + time_delta
calc_yyyymmdd = calc_datetime.strftime("%Y%m%d")
calc_hhmm = calc_datetime.strftime("%H%M")
line_val = np.float64(row[2])
#convert units by line (only if value is >= than 0
if line_val >= 0:
if (unit.lower() != 'ppb') & (unit.lower() != 'ppbv'):
if unit == 'ug/m3':
#print 'converting units, temp = 20degC'
#calculate conversion factor from mg/m3 assuming 20 degC and 1 atm - default for O3 instruments
#R/MW*(TEMP0C(K)*TEMP(degC)/P(hPa)/10
conv_fact = 8.3144/mol_mass*(273.15+20)/(1013.25/10)
line_val = conv_fact*line_val
#print 'Converting Units from ug/m3 20degC to ppbv'
elif unit == 'ug_N/m3':
conv_fact = 8.3144/mol_mass*(273.15+20)/(1013.25/10)
line_val = conv_fact*line_val
#print 'Converting Units from ug/Nm3 20degC to ppbv'
elif (unit == 'ppm') or (unit == 'ppmv'):
line_val = line_val*1e3
#print 'Converting Units from ppmv to ppbv'
elif (unit == 'ppt') or (unit == 'pptv'):
line_val = line_val/1e3
#print 'Converting Units from pptv to ppbv'
else:
print 'Unknown Unit'
data_valid = False
1+'a'
if file_res == '1h':
yyyymmdd=np.append(yyyymmdd,calc_yyyymmdd)
hhmm=np.append(hhmm,calc_hhmm)
vals = np.append(vals,line_val)
flags = np.append(flags,np.float64(row[3]))
elif file_res == '1d':
yyyymmdd=np.append(yyyymmdd,calc_yyyymmdd)
hhmm=np.append(hhmm,'0000')
vals = np.append(vals,line_val)
flags = np.append(flags,np.float64(row[3]))
for j in range(1,24):
time_delta = datetime.timedelta(days = days_since_start,hours = j)
calc_datetime = start_datetime + time_delta
vals = np.append(vals,vals[-1])
flags = np.append(flags,flags[-1])
yyyymmdd = np.append(yyyymmdd,calc_datetime.strftime("%Y%m%d"))
hhmm = np.append(hhmm,calc_datetime.strftime("%H%M"))
elif file_res == '1mo':
yyyymmdd=np.append(yyyymmdd,calc_yyyymmdd)
hhmm=np.append(hhmm,'0000')
vals = np.append(vals,line_val)
flags = np.append(flags,np.float64(row[3]))
month_days = monthrange(int(yyyymmdd[-1][:4]), int(yyyymmdd[-1][4:6]))[1]
for j in range(1,24*month_days):
time_delta = datetime.timedelta(days = days_since_start,hours = j)
calc_datetime = start_datetime + time_delta
vals = np.append(vals,vals[-1])
flags = np.append(flags,flags[-1])
yyyymmdd = np.append(yyyymmdd,calc_datetime.strftime("%Y%m%d"))
hhmm = np.append(hhmm,calc_datetime.strftime("%H%M"))
if row[0] == 'starttime':
start_read = True
count+=1
if (y == year_array[-1]):
#convert all invalids by flags to -99999
test_inv = flags != 0
if len(test_inv) != 0:
vals[test_inv] = -99999
#any values less than zero are -99999
test_inv = vals < 0
if len(test_inv) != 0:
vals[test_inv] = -99999
#do additional invalid test, as flags not always correct
#test_inv_2 = vals > 300
#vals[test_inv_2] = -99999
#put o3 vals into full grid
date_con = np.array(yyyymmdd).astype(int)
time_con = np.array(hhmm).astype(int)
#find matching times between actual times and grid of times, return big array of indices of matched indices in grid
converted_time = date_process(date_con,time_con,start_year)
converted_time = np.round(converted_time,decimals=5)
syn_grid_time = np.arange(0,n_days,1./24)
syn_grid_time = np.round(syn_grid_time,decimals=5)
#find matching times between actual times and grid of times, return big array of indices of matched indices in grid
indices = np.searchsorted(syn_grid_time, converted_time, side='left')
vals = np.array(vals)
#make sure no data is past end year
index_test = indices < len(full_data)
indices = indices[index_test]
vals = vals[index_test]
full_data[indices] = vals
#get mode of metadata
lat = np.float64(stats.mode(all_latitudes)[0][0])
lon = np.float64(stats.mode(all_longitudes)[0][0])
alt = np.float64(stats.mode(all_altitudes)[0][0])
mm = stats.mode(all_mm)[0][0]
#check site is not urban using anthrome map from 2000
anthfile = '/work/home/db876/plotting_tools/core_tools/anthro2_a2000.nc'
anthload = Dataset(anthfile)
class_valid,anthrome_class_name = modules.anthrome_classify(anthload,[lat],[lon])
if class_valid == 'invalid':
data_valid = False
print 'Site Invalid, site classed as urban by anthrome map.'
#get measurement method
if (mm == 'uv_abs') or (mm == 'chemiluminesc') or (mm == 'uv_fluoresc'):
if species == 'O3':
mm = 'ultraviolet photometry'
if (species == 'NO') or (species == 'NO2') or (species == 'CO'):
mm = 'chemiluminescence'
elif (mm == 'ndir') or (mm == 'infrared_absorption'):
mm = 'non-dispersive infrared spectroscopy'
elif (mm == 'GC-HgO'):
mm = 'gas chromatography reduction gas detection'
elif (mm == 'tracegas_monitor'):
mm = 'cavity attenuated phase shift spectroscopy'
elif (mm == 'filter_1pack') or (mm == 'filter_2pack') or (mm == 'filter_3pack'):
if species == 'NO2':
mm = 'griess saltzman colorimetric'
elif species == 'CO':
mm = 'ion chromatography'
elif (mm == 'steel_canister'):
mm = 'gas chromatography flame ionisation detection'
elif (mm == 'online_gc'):
mm = 'online gas chromatography'
elif (mm == 'glass_sinter') or (mm == 'abs_solution') or (mm == 'filter_abs_solution') or (mm == 'abs_tube') or (mm == 'continuous_colorimetric'):
mm = 'griess saltzman colorimetric'
elif (mm == 'NaJ_solution'):
mm = 'flame ionisation detection'
elif (mm == 'doas'):
mm = 'differential optical absorption spectrosocopy'
elif (mm == 'diffusion_tube'):
mm = 'diffusive sampler'
elif (mm == 'NA') or (mm == ''):
if species == 'O3':
mm = 'ultraviolet photometry'
if species == 'CO':
mm = 'non-dispersive infrared spectroscopy'
if species == 'NO2':
mm = 'chemiluminescence'
if species == 'NO':
mm = 'chemiluminescence'
if species == 'ISOP':
mm = 'gas chromatography flame ionisation detection'
else:
print mm
1+'a'
#do data quality checks
full_data,data_valid = modules.quality_check(full_data,data_valid,data_resolution,alt,grid_dates,start_year,end_year)
#convert file res to standard format
if file_res == '1h':
file_res = 'H'
elif file_res == '1d':
file_res = 'D'
elif file_res == '1mo':
file_res = 'M'
#no raw class so set as na
raw_class_name = 'na'
#set sampling as average
st = 'average'
return c,full_data,data_valid,lat,lon,alt,raw_class_name,anthrome_class_name,mm,st,file_res
def site_iter_process(valid_refs,c):
#for site_ref in valid_refs:
site_ref = valid_refs[c]
data_valid = True
print 'ref = ',site_ref
site_test = all_refs == site_ref
site_yyyymmdd = yyyymmdd[site_test]
site_hhmm = hhmm[site_test]
site_vals = vals[site_test]
site_vals = np.float64(site_vals)
#convert all invalids to -99999
test_inv = site_vals < 0
site_vals[test_inv] = -99999
#put vals into full grid
date_con = np.array(site_yyyymmdd).astype(int)
time_con = np.array(site_hhmm).astype(int)
#create max possible o3 grid
full_data = np.empty(n_hours)
full_data[:] = -99999
#find matching times between actual times and grid of times, return big array of indices of matched indices in grid
converted_time = modules.date_process(date_con,time_con,start_year)
converted_time = np.round(converted_time,decimals=5)
syn_grid_time = np.arange(0,n_days,1./24)
syn_grid_time = np.round(syn_grid_time,decimals=5)
#find matching times between actual times and grid of times, return big array of indices of matched indices in grid
indices = np.searchsorted(syn_grid_time, converted_time, side='left')
site_vals = np.array(site_vals)
full_data[indices] = site_vals
meta_index = meta_refs.index(site_ref)
tz = float(meta_tz[meta_index])
lat = np.float64(meta_lats[meta_index])
lon = np.float64(meta_lons[meta_index])
alt = np.float64(meta_alts[meta_index])
raw_class_name = meta_class[meta_index]
anthrome_class_name = class_name[meta_index]
#check tz is whole number else skip site
if (tz % 1) != 0:
data_valid = False
print 'Timezone is not a whole number. Skipping.'
#correct timezone to UTC
tz = int(tz)
if tz < 0:
#get rid of values at start and append -99999's at end
cut = full_data[:tz]
for num in range(np.abs(tz)):
cut = np.insert(cut,0, -99999)
full_data = cut
elif tz > 0:
#put -99999's at start and get rid of values at end
cut = full_data[tz:]
for num in range(tz):
cut = np.append(cut, -99999)
full_data = cut
#if species is CO then convert units from ppmv to ppbv
if species == 'CO':
valid_inds = full_data != -99999
full_data[valid_inds] = full_data[valid_inds]*1e3
#do data quality checks
full_data,data_valid = modules.quality_check(full_data,data_valid,data_resolution,alt,grid_dates,start_year,end_year)
#set sampling as average
if (species == 'O3') or (species == 'CO') or(species == 'NO') or (species == 'NO2'):
st = 'average'
elif (species == 'ISOP'):
st = 'flask'
#set site file resolution
if (species == 'O3') or (species == 'CO') or(species == 'NO') or (species == 'NO2'):
file_res = 'H'
elif (species == 'ISOP'):
file_res = 'D'
#check file res is ok for output res
if (output_res == 'H'):
if (file_res == 'D') or (file_res == 'M'):
print 'File resolution has to be Minimum Hourly. Skipping'
data_valid = False
return c,full_data,data_valid,-999,-999,-999,'na','na','na','na','na'
elif (output_res == 'D'):
if (file_res == 'M'):
print 'File resolution has to be Minimum Daily. Skipping'
data_valid = False
return c,full_data,data_valid,-999,-999,-999,'na','na','na','na','na'
#set mm
if species == 'O3':
mm = 'ultraviolet photometry'
elif (species == 'NO') or (species == 'NO2'):
mm = 'chemiluminescence'
elif species == 'CO':
mm = 'non-dispersive infrared spectrometry'
elif species == 'ISOP':
mm = 'gas chromatography flame ionisation detection'
return c,full_data,data_valid,lat,lon,alt,raw_class_name,anthrome_class_name,mm,st,file_res
def site_iter_process(valid_refs,c):
#for ref_i in range(len(valid_refs)):
data_valid = True
site_ref = valid_refs[c]
print 'Current Ref is = ', site_ref
s_files = glob.glob('/work/home/db876/observations/surface/%s/CAPMON/ozon_smpls_%s*'%(species,site_ref))
site_files = []
for y in year_array:
for f in s_files:
if str(y) in f:
site_files.append(f)
site_files = modules.natsorted(site_files)
yymmdd = []
hhmm = []
vals = []
#create max possible o3 grid
full_data = np.empty(n_hours)
full_data[:] = -99999
for file_i in range(len(site_files)):
count = 0
meta_start = -99999
start_read_1 = False
start_read_2 = False
with open(site_files[file_i], 'rb') as f:
reader = csv.reader(f,delimiter=',')
print site_files[file_i]
for row in reader:
#print count
#break out of loop at bottom of file
if (start_read_2 == True) & (row[0] == '*TABLE ENDS'):
break
#get metadata
try:
if (row[0] =='*TABLE NAME') & (row[1] == 'Site information'):
meta_start = count+2
except:
pass
if count == meta_start:
lat_i = row.index('Latitude: decimal degrees')
lon_i = row.index('Longitude: decimal degrees')
try:
alt_i = row.index('Ground elevation: above mean sea level')
except:
alt_i = row.index('Ground altitude')
class_i = row.index('Site land use')
if count == (meta_start+6):
latitude = row[lat_i]
longitude = row[lon_i]
altitude = row[alt_i]
raw_class_name = row[class_i]
#get data
if start_read_2 == True:
#read dates, times, and vals
date = row[8]
time = row[9]
yymmdd.append(date[:4]+date[5:7] + date[8:])
hhmm.append(time[:2]+time[3:])
quality_code = row[13]
if quality_code == 'V0':
vals = np.append(vals,np.float64(row[12]))
else:
vals = np.append(vals,-99999)
try:
if (row[0] == '*TABLE NAME') & (row[1] == 'OZONE_HOURLY'):
start_read_1 = True
except:
pass
if (start_read_1 == True) & (row[0] == '*TABLE COLUMN UNITS'):
unit = row[12]
if (start_read_1 == True) & (row[0] == '*TABLE BEGINS'):
start_read_2 = True
count+=1
#convert all invalids to -99999
test_inv = vals < 0
vals[test_inv] = -99999
#put o3 vals into full grid
date_con = np.array(yymmdd).astype(int)
time_con = np.array(hhmm).astype(int)
#find matching times between actual times and grid of times, return big array of indices of matched indices in grid
converted_time = modules.date_process(date_con,time_con,start_year)
converted_time = np.round(converted_time,decimals=5)
syn_grid_time = np.arange(0,n_days,1./24)
syn_grid_time = np.round(syn_grid_time,decimals=5)
#find matching times between actual times and grid of times, return big array of indices of matched indices in grid
indices = np.searchsorted(syn_grid_time, converted_time, side='left')
vals = np.array(vals)
#make sure no data is past end year
index_test = indices < len(full_data)
indices = indices[index_test]
vals = vals[index_test]
full_data[indices] = vals
#get metadata
lat = np.float64(latitude)
lon = np.float64(longitude)
alt = np.float64(altitude)
#check site is valid by class
if ('Urban' in raw_class_name) or ('urban' in raw_class_name):
data_valid=False
print 'Data is invalid. Raw Class is Urban.'
#check site is not urban using anthrome map from 2000
anthfile = '/work/home/db876/plotting_tools/core_tools/anthro2_a2000.nc'
anthload = Dataset(anthfile)
class_result,anthrome_class_name = modules.anthrome_classify(anthload,[lat],[lon])
if class_result == 'invalid':
data_valid = False
print 'Site Invalid, site classed as urban by anthrome map.'
#do data quality checks
full_data,data_valid = modules.quality_check(full_data,data_valid,data_resolution,np.float64(altitude),grid_dates,start_year,end_year)
#set measurement method
mm = 'ultraviolet photometry'
#set site file resolution
file_res = 'H'
#set sampling as average
st = 'average'
return c,full_data,data_valid,lat,lon,alt,raw_class_name,anthrome_class_name,mm,st,file_res
def site_iter_process(valid_refs, c):
#process data for each site at a time
#for site_ref in valid_refs:
site_ref = valid_refs[c]
data_valid = True
print 'ref = ', site_ref
site_test = all_refs == site_ref
site_yyyymmdd = yyyymmdd[site_test]
site_hhmm = hhmm[site_test]
site_vals = vals[site_test]
site_mm = all_mm[site_test]
site_units = units[site_test]
if species == 'ISOP':
site_sample_len = sample_len[site_test]
#check for data below limit of detection (only for ISOP) as other species have LOD check by line in file. If it is change to -99999
#LOD for ISOP if 0.01 ppbv
if species == 'ISOP':
lod_test = site_vals < 0.01
#convert from ppm to ppb
if (species == 'O3') or (species == 'NO') or (species == 'NO2'):
for i in range(len(site_vals)):
if site_units[i] == 'Parts per million':
site_vals[i] = site_vals[i] * 1.e3
elif site_units[i] == 'Parts per billion':
pass
else:
print site_units[i]
1 + 'a'
# convert from ppbC to ppb
if species == 'ISOP':
for i in range(len(site_vals)):
#078 is Parts per billion Carbon, Isoprene has 5 Carbons
if site_units[i] == '078':
site_vals[i] = site_vals[i] / 5.
#008 is Parts per billion
if site_units[i] == '008':
pass
#101 is Parts per million Carbon
if site_units[i] == '101':
site_vals[i] = (site_vals[i] / 5.) * 1.e3
site_vals[lod_test] = -99999
#put vals into full grid
date_con = np.array(site_yyyymmdd).astype(int)
time_con = np.array(site_hhmm).astype(int)
#create max possible o3 grid
full_data = np.empty(n_hours)
full_data[:] = -99999
#find matching times between actual times and grid of times, return big array of indices of matched indices in grid
converted_time = modules.date_process(date_con, time_con, start_year)
converted_time = np.round(converted_time, decimals=5)
syn_grid_time = np.arange(0, n_days, 1. / 24)
syn_grid_time = np.round(syn_grid_time, decimals=5)
#find matching times between actual times and grid of times, return big array of indices of matched indices in grid
indices = np.searchsorted(syn_grid_time, converted_time, side='left')
site_vals = np.array(site_vals)
#if date goes past where it should finish, omit it.
inv_i = indices < len(full_data)
indices = indices[inv_i]
site_vals = site_vals[inv_i]
full_data[indices] = site_vals
#get site meta
meta_index = meta_refs.index(site_ref)
lat = np.float64(meta_lats[meta_index])
lon = np.float64(meta_lons[meta_index])
alt = np.float64(meta_alts[meta_index])
raw_class_name = meta_class[meta_index]
anthrome_class_name = class_name[meta_index]
#get measurement method, take mode of big methods array
site_mm = stats.mode(site_mm)[0][0]
if (site_mm.upper() == 'INSTRUMENTAL-ULTRAVIOLETABSORPTION') or (
site_mm.upper() == 'INSTRUMENTAL-ULTRAVIOLET2BMODEL202'
) or (site_mm.upper() == 'INSTRUMENTAL-UVPHOTOMETRIC') or (
site_mm.upper() == 'INSTRUMENTAL-ULTRAVIOLETRADIATIONABSORBTN'
) or (site_mm.upper() == 'INSTRUMENTAL-ULTRAVIOLET') or (
site_mm.upper() == 'INSTRUMENTAL-ULTRAVIOLETPHOTOMETRY') or (
site_mm.upper() ==
'INSTRUMENTAL-UVABSORPTIONPHOTOMETRY/UV2BMODEL202AND205') or (
site_mm.upper() == 'INSTRUMENTAL-ECOTECHSERINUS10'):
mm = 'ultraviolet photometry'
elif (site_mm.upper() == 'INSTRUMENTAL-CHEMILUMINESCENCE') or (
site_mm.upper() == 'INSTRUMENTAL-GASPHASECHEMILUMINESCENCE'
) or (site_mm.upper(
) == 'INSTRUMENTAL-CHEMILUMINESCENCEAPIMODEL265EANDT265') or (
site_mm.upper() ==
'LOWLEVELNOXINSTRUMENTAL-TECO42SCHEMILUMINESCENCE'
) or (site_mm.upper() == 'INSTRUMENTAL-GAS-PHASECHEMILUMINESCENCE') or (
site_mm.upper() ==
'INSTRUMENTAL-CHEMILUMINESCENCETELEDYNEAPIT200UPPHOTOLYTIC'
) or (site_mm.upper(
) == 'INSTRUMENTAL-CHEMILUMINESCENCETELEDYNEAPI200EU/501') or (
site_mm.upper() == 'INSTRUMENTAL-CHEMILUMINESCENCEECOTECHEC9841T'
) or (site_mm.upper(
) == 'TELEDYNE-APIMODEL200EUPORT200UP-PHOTOLYTIC-CHEMILUMINESCENCE') or (
site_mm.upper() ==
'INSTRUMENTAL-CHEMILUMINESCENCETHERMOELECTRON42C-TL,42I-TL'
) or (site_mm.upper(
) == 'INSTRUMENTAL-CHEMILUMINESCENCERHODAMINEBDYE') or (site_mm.upper(
) == 'INSTRUMENTAL-CHEMILUMINESCENCETHERMOELECTRON42C-Y,42I-Y') or (
site_mm.upper() == 'INSTRUMENTAL-CHEMILUMINESCENCEECOTECHEC9843'):
mm = 'chemiluminescence'
elif (site_mm.upper() == 'INSTRUMENTAL-OPENPATHO3ANALYZER') or (
site_mm.upper() == 'INSTRUMENTAL-OPENPATHNOANALYZER'):
mm = 'differential optical absorption spectrosocopy'
elif (site_mm.upper() ==
'TELEDYNEMODELT500U-CAVITYATTENUATEDPHASESHIFTSPECTROSCOPY'):
mm = 'cavity attenuated phase shift spectroscopy'
elif (site_mm.upper() == 'INSTRUMENTAL-COLORIMETRIC-GRIESS-SALTZMAN') or (
site_mm.upper() == 'INSTRUMENTAL-COLORIMETRIC'):
mm = 'griess saltzman colorimetric'
elif (site_mm.upper() == 'INSTRUMENTAL-COLORIMETRIC-LYSHKOW(MOD)'):
mm = 'lyshkow colorimetric '
elif (site_mm.upper() == 'INSTRUMENTAL-COULOMETRIC'):
mm = 'coulometry'
else:
print site_mm.upper()
1 + 'a'
#do data quality checks
full_data, data_valid = modules.quality_check(
full_data, data_valid, data_resolution, alt, grid_dates, start_year,
end_year)
#set site file resolution
file_res = 'H'
#set sampling as average
st = 'average'
return c, full_data, data_valid, lat, lon, alt, raw_class_name, anthrome_class_name, mm, st, file_res