Convert Naryn monthly flow rate data to monthly volumes.

Assumes the flow rate is an average that can be multiplied by days in the month for total monthly volume.



In [ ]:

    
from __future__ import print_function
import calendar
from charistools.hypsometry import Hypsometry
from charistools.timeSeries import TimeSeries
import datetime as dt
import pandas as pd
import numpy as np
import os
import re
from time import strptime



In [ ]:

    
%cd /Users/brodzik/projects/CHARIS/streamflow/From_Alice_Naryn
%ls



In [ ]:

    
file = "Kyrygzstan_Naryn_flow_monthly_average_1933-2015.xlsx"



In [ ]:

    
# Use sheetname if pandas < 0.21.0
# use sheet_name if later
# Returns dict of sheets in file
sheets = pd.read_excel(io=file, sheetname=None)



In [ ]:

    
for key in sheets:
    print("key=%s" % key)



In [ ]:

    
# Set the source name from the original file (controls which sheet is extracted)
# and the CHARIS-format drainageID
sheetName = "Naryn town gauge (Naryn Darya)"
drainageID = "SY_Naryn_at_NarynTown"



In [ ]:

    
df = sheets[sheetName]
df



In [ ]:

    
description = df.columns[0]
description



In [ ]:

    
geolocation = df.iloc[0][0]
geolocation



In [ ]:

    
units = df.iloc[1][0]
units



In [ ]:

    
description, geolocation, units



In [ ]:

    
def dmsds_to_decdeg(deg, min, sec, decsec):
    return float(deg) + float(min)/60. + (float(sec) + float(decsec)/100.)/60./60.



In [ ]:

    
# Convert lat/lon positions to decimal degrees
p = re.compile(r'latitude:(.*)longitude:(.*)\(')
m = p.search(geolocation)
if m :
    print(m.group(1))
    print(m.group(2))
    
    (deg, min, sec, decsec) = m.group(1).split()
    latitude = dmsds_to_decdeg(deg, min, sec, decsec)
    (deg, min, sec, decsec) = m.group(2).split()
    longitude = dmsds_to_decdeg(deg, min, sec, decsec)
    print("lat/lon: %s/%s" % (latitude, longitude))
    
else:
    print("no match")



In [ ]:

    
# Drop the first 2 data rows, they are just metadata we have already scraped
df = df.drop([0,1])



In [ ]:

    
#sheets['Naryn town gauge (Naryn Darya)']
df.loc[2].values



In [ ]:

    
# Move the contents of data row 2 into the column headers
df.columns = df.loc[2].values
df



In [ ]:

    
# Now drop the data row 2
df = df.drop([2])
df



In [ ]:

    
# Move the df index to the Year column
df = df.set_index('Year')
df



In [ ]:

    
df['Jan'][1933]



In [ ]:

    
# Write the result as a hypsometry for ti-melt modeling calibration steps, this needs to look like this:
# # Station ID: 35749301
# # Basin: Hunza at Dainyor Bridge
# # Latitude: 35.92778
# # Longitude: 74.37639
# # Elevation: 1370 m
# # Area: 13157 km**2
# # Units: km**3
# # COLUMNS: year month runoff
# 1966  1      0.15
# 1966  2      0.12
# 1966  3      0.10
outfile = "%s.month_runoff.dat" % drainageID
outfile



In [ ]:

    
%pwd



In [ ]:

    
comments=["Description: Derived runoff volume from monthly flow rates",
          "Flow rate file: %s" % os.path.basename(file),
          "Basin: %s" % drainageID,
          "Latitude: %.6f" % latitude,
          "Longitude: %.6f" % longitude,
          "Elevation: unknown",
          "Area: unknown",
          "Units: km**3",
          "COLUMNS: year month runoff"]
f = open(outfile, "w")
for line in comments:
    f.write("# %s\n" % line)

Convert the monthly average flow rates to volumes using:

$Volume(km^3/month) = \frac{m^3}{s} * \frac{1 km}{1000 m} * \frac{1 km}{1000 m}* \frac{1 km}{1000 m} * \frac{60 * 60 * 24 s}{1 day} * \frac{N days}{month}$



In [ ]:

    
m_per_km = 1000.
s_per_day = 60. * 60. * 24.
conversion_factor = s_per_day / (m_per_km * m_per_km * m_per_km)
for year in df.index:
    for mmm in df.columns[1:13]:
        month = strptime(mmm,'%b').tm_mon
        days_in_month = calendar.monthrange(year, int(month))[1]
        volume_km3 = df[mmm][year] * conversion_factor * days_in_month
        print("%4d %d %d %f %f" % (year, month, days_in_month, df[mmm][year], volume_km3))
        print("%4d %2d %.6f" % (year, month, volume_km3),
              file=f)



In [ ]:

    
f.close()



In [ ]:

    
%ls *month_runoff*



In [ ]:

    
%pwd
ts = TimeSeries("../SY_Naryn_at_NarynTown.month_runoff.dat")



In [ ]:

    
ts.comments

Convert monthly runoff to annual runoff



In [ ]:

    
def convert_monthly_to_annual_runoff(monthFile):
    monthTS = TimeSeries(monthFile)
    # Add new description to beginning
    monthTS.comments.insert(0, "# Monthly file: %s" % monthFile)
    monthTS.comments.insert(0, "# Description: Annual runoff volume, created %s" % str(dt.datetime.now()))
    # And replace Column descriptor line:
    monthTS.comments = monthTS.comments[:-1]
    monthTS.comments.extend(["# COLUMNS: year runoff"])
        
    #print(monthTS.data)
    monthTS.data["year"] = monthTS.data.index.year
    annual = monthTS.data.groupby(["year"])["runoff"].sum()
    
    # Make the new outfile by replacing "monthly" in monthFile
    p = re.compile("month")
    annualFile = p.sub("annual", monthFile)
    
    f = open(annualFile, "w")
    for line in monthTS.comments:
        f.write("%s\n" % line)
        
    for key, value in annual.iteritems():
        print("%4d %.6f" % (key, value), file=f)
        
    f.close()



In [ ]:

    
%cd /Users/brodzik/projects/CHARIS/streamflow/From_Alice_Naryn
%ls



In [ ]:

    
monthFile = "SY_Naryn_at_NarynTown.month_runoff.dat"



In [ ]:

    
convert_monthly_to_annual_runoff(monthFile)



In [ ]:

    
%pwd



In [ ]:

    
annualFile = "SY_Naryn_at_NarynTown.annual_runoff.dat"
new = TimeSeries(annualFile)
print(new.data)
print(new.comments)
print(new.units)



In [ ]:

    
%cat $annualFile



In [ ]:

    
line = "# Units: km**3"
line.strip().lstrip('# Units: ')



In [ ]: