Coweeta Hydrologic Laboratory has been collecting climate data since it was established by the US Forest Service in 1934¶

The data available online through the Global Historical Climatology Network includes data from Denver Co & Coweeta which will be compared in this post.¶

  • The Global Historical Climatology Network daily data integrates data from 30 different sources
  • Observations are from a variety of meteorological organizations using land-based stations around the globe.
  • Temperatures can be displayed in degrees Fahrenheit, but are collected in degrees Celsius
  • There are three steps used to integrate data from different countries. Some locations contain historic data only and a quality assurance protocol is used although the data is not homogenous.

Sources:¶

  • Menne, Matthew J., Imke Durre, Bryant Korzeniewski, Shelley McNeill, Kristy Thomas, Xungang Yin, Steven Anthony, Ron Ray, Russell S. Vose, Byron E.Gleason, and Tamara G. Houston (2012): Global Historical Climatology Network - Daily (GHCN-Daily), Version 3. Coweeta Hydrologic Station USC00312102. NOAA National Climatic Data Center. doi:10.7289/V5D21VHZ Accessed: Nov. 18, 2024.
  • Matthew J. Menne, Imke Durre, Russell S. Vose, Byron E. Gleason, and Tamara G. Houston, 2012: An Overview of the Global Historical Climatology Network-Daily Database. J. Atmos. Oceanic Technol., 29, 897-910. doi:10.1175/JTECH-D-11-00103.1.
  • https://www.ncei.noaa.gov/pub/data/ghcn/daily/readme.txt
In [39]:
# Import required packages
import holoviews as hv
import hvplot.pandas
import matplotlib.pyplot as plt
import pandas as pd
In [2]:
# Common statistical plots for tabular data
import seaborn as sns
# Fit an OLS linear regression
from sklearn.linear_model import LinearRegression
In [23]:
GHCnd_url = ('https://www.ncei.noaa.gov/access/services/data/v1'
'?dataset=daily-summaries'
'&dataTypes=TOBS,PRCP'
'&stations=USC00312102'
'&startDate=1961-01-01'
'&endDate=2024-09-30'
'&units=standard'
)
GHCnd_url
Out[23]:
'https://www.ncei.noaa.gov/access/services/data/v1?dataset=daily-summaries&dataTypes=TOBS,PRCP&stations=USC00312102&startDate=1961-01-01&endDate=2024-09-30&units=standard'
In [24]:
# Import data into Python from NCEI API
nc_climate_df = pd.read_csv(GHCnd_url)

nc_climate_df
Out[24]:
STATION DATE PRCP TOBS
0 USC00312102 1961-01-01 1.12 39.0
1 USC00312102 1961-01-02 0.00 23.0
2 USC00312102 1961-01-03 0.00 29.0
3 USC00312102 1961-01-04 0.00 33.0
4 USC00312102 1961-01-05 0.00 21.0
... ... ... ... ...
23278 USC00312102 2024-09-26 3.48 63.0
23279 USC00312102 2024-09-27 5.61 70.0
23280 USC00312102 2024-09-28 0.60 61.0
23281 USC00312102 2024-09-29 0.27 61.0
23282 USC00312102 2024-09-30 0.00 62.0

23283 rows × 4 columns

In [33]:
# Download the climate data
nc_climate_df = pd.read_csv(
    GHCnd_url,
    index_col='DATE',
    parse_dates=True,
    na_values=['NaN'],
    )
nc_climate_df.dropna()

nc_climate_df
Out[33]:
STATION PRCP TOBS
DATE
1961-01-01 USC00312102 1.12 39.0
1961-01-02 USC00312102 0.00 23.0
1961-01-03 USC00312102 0.00 29.0
1961-01-04 USC00312102 0.00 33.0
1961-01-05 USC00312102 0.00 21.0
... ... ... ...
2024-09-26 USC00312102 3.48 63.0
2024-09-27 USC00312102 5.61 70.0
2024-09-28 USC00312102 0.60 61.0
2024-09-29 USC00312102 0.27 61.0
2024-09-30 USC00312102 0.00 62.0

23283 rows × 3 columns

In [34]:
if not isinstance(nc_climate_df.index, pd.DatetimeIndex):
    # Assuming a column 'Date' or similar contains datetime information
    nc_climate_df.set_index('Date', inplace=True)
In [35]:
nc_climate_df['temp_c'] = (nc_climate_df['TOBS'] - 32) * 5 / 9

nc_climate_df
Out[35]:
STATION PRCP TOBS temp_c
DATE
1961-01-01 USC00312102 1.12 39.0 3.888889
1961-01-02 USC00312102 0.00 23.0 -5.000000
1961-01-03 USC00312102 0.00 29.0 -1.666667
1961-01-04 USC00312102 0.00 33.0 0.555556
1961-01-05 USC00312102 0.00 21.0 -6.111111
... ... ... ... ...
2024-09-26 USC00312102 3.48 63.0 17.222222
2024-09-27 USC00312102 5.61 70.0 21.111111
2024-09-28 USC00312102 0.60 61.0 16.111111
2024-09-29 USC00312102 0.27 61.0 16.111111
2024-09-30 USC00312102 0.00 62.0 16.666667

23283 rows × 4 columns

In [36]:
# Plot the data using .plot
nc_climate_df.plot(
    y='temp_c',
    title='Daily Temperature at Coweeta Hydrological Lab',
    xlabel='Date',
    ylabel='Temperature ($^\circ$F)')
Out[36]:
<Axes: title={'center': 'Daily Temperature at Coweeta Hydrological Lab'}, xlabel='Date', ylabel='Temperature ($^\\circ$F)'>
No description has been provided for this image
In [41]:
# Calculate the mean temperature
mean_temp = nc_climate_df['temp_c'].mean()
print("Mean Temperature:", mean_temp)

Mean Temperature: 10.085713600498837
In [43]:
# Group by year and calculate the mean temperature
yearly_avg_temp = nc_climate_df.groupby(nc_climate_df.index.year)['temp_c'].mean()

print(yearly_avg_temp)

DATE
1961    10.856227
1962    11.321157
1963    10.462454
1964    13.897243
1965    11.004566
          ...    
2020    10.918336
2021    10.313546
2022    10.111111
2023    10.490107
2024    12.862936
Name: temp_c, Length: 64, dtype: float64
In [69]:
# Convert the Series to a DataFrame with 'Year' as the index
nc_ann_climate_df = yearly_avg_temp.to_frame(name='Average_Temperature')

# Reset the index to make 'Year' a regular column
nc_ann_climate_df.reset_index('DATE', inplace=True)
print(nc_ann_climate_df.head())

   DATE  Average_Temperature
0  1961            10.856227
1  1962            11.321157
2  1963            10.462454
3  1964            13.897243
4  1965            11.004566
In [67]:
nc_ann_climate_df = nc_ann_climate_df.loc['1961':'2024']

# Create a Linear Regression model
model = LinearRegression()

# Fit the model on the training data
model.fit(nc_ann_climate_df[['DATE']], nc_ann_climate_df['Average_Temperature'])

# Calculate and print the metrics
slope = model.coef_[0]
print(f"Slope: {slope:.2f} degrees Celsius per year")

Slope: 0.00 degrees Celsius per year
In [71]:
ax = sns.regplot(
    x=nc_ann_climate_df.DATE, 
    y=nc_ann_climate_df.Average_Temperature,
    color='red',
    line_kws={'color': 'black'})
ax.set(
    title='Annual Average Daily Temperature over time at Coweeta Lab',
    xlabel='Year',
    ylabel='Temperature ($^\circ$C)'
)
plt.show()
No description has been provided for this image