NRT data

import time

import matplotlib.pyplot as plt
import pandas as pd

import uscrn

Show code cell output

Hide code cell output

Matplotlib is building the font cache; this may take a moment.

Recent hourly data

With uscrn.get_nrt_data(), we can load recent data (near-real-time) from USCRN by specifying the period we want and from which dataset. Here, we request the 6 most recent files.

now = pd.Timestamp.now("UTC")
print(now)

df = uscrn.get_nrt_data(
    (-6, None),
    "hourly",
    n_jobs=2,
)

Show code cell output

Hide code cell output

2026-06-02 02:50:26.998775+00:00

Discovering files...

  Looking for files in these years
  - 2026

Found 6 file(s) to load
https://www.ncei.noaa.gov/pub/data/uscrn/products/hourly02/updates/2026/CRN60H0203-202606012000.txt
...
https://www.ncei.noaa.gov/pub/data/uscrn/products/hourly02/updates/2026/CRN60H0203-202606020100.txt
Reading files...

[Parallel(n_jobs=2)]: Using backend LokyBackend with 2 concurrent workers.

[Parallel(n_jobs=2)]: Done   1 tasks      | elapsed:    1.5s

[Parallel(n_jobs=2)]: Done   4 out of   6 | elapsed:    1.8s remaining:    0.9s

[Parallel(n_jobs=2)]: Done   6 out of   6 | elapsed:    2.0s finished

df

	wban	utc_time	lst_time	crx_vn	longitude	latitude	t_calc	t_hr_avg	t_max	t_min	...	soil_moisture_5	soil_moisture_10	soil_moisture_20	soil_moisture_50	soil_moisture_100	soil_temp_5	soil_temp_10	soil_temp_20	soil_temp_50	soil_temp_100
0	03047	2026-06-01 19:00:00	2026-06-01 13:00:00	2.622	-102.809998	31.620001	32.700001	32.299999	33.099998	31.200001	...	0.037	0.039	0.051	0.066	0.062	40.599998	32.500000	30.700001	28.799999	28.400000
1	03048	2026-06-01 19:00:00	2026-06-01 12:00:00	2.622	-106.889999	34.360001	32.000000	31.000000	32.400002	29.500000	...	0.047	0.059	0.059	0.059	0.066	40.099998	30.299999	26.600000	25.700001	24.799999
2	03054	2026-06-01 19:00:00	2026-06-01 13:00:00	2.622	-102.769997	33.959999	32.099998	32.099998	33.000000	31.400000	...	0.168	0.198	0.152	0.108	0.133	31.600000	26.100000	23.799999	23.700001	20.900000
3	03055	2026-06-01 19:00:00	2026-06-01 13:00:00	2.622	-101.589996	36.599998	32.799999	32.299999	32.900002	31.500000	...	0.201	0.234	0.200	0.232	0.160	24.400000	22.400000	20.400000	19.900000	19.299999
4	03060	2026-06-01 19:00:00	2026-06-01 12:00:00	2.622	-107.690002	38.540001	20.799999	20.299999	21.100000	19.500000	...	0.080	0.144	0.194	0.258	0.199	23.799999	17.500000	15.200000	14.500000	12.100000
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
1372	96404	2026-06-02 00:00:00	2026-06-01 15:00:00	2.514	-141.210007	62.740002	15.700000	17.500000	18.700001	15.500000	...	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN
1373	96405	2026-06-02 00:00:00	2026-06-01 15:00:00	2.514	-145.350006	60.470001	13.600000	13.600000	14.100000	13.300000	...	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN
1374	96407	2026-06-02 00:00:00	2026-06-01 15:00:00	2.515	-159.000000	66.559998	13.700000	13.700000	14.300000	12.600000	...	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN
1375	96408	2026-06-02 00:00:00	2026-06-01 15:00:00	2.515	-150.869995	63.450001	12.300000	12.200000	12.600000	11.900000	...	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN
1376	96409	2026-06-02 00:00:00	2026-06-01 15:00:00	2.514	-149.399994	68.650002	6.300000	6.600000	8.300000	5.500000	...	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN

1377 rows × 36 columns

Sometimes, a few sites can have times an hour earlier than the others. See the notes in uscrn.get_nrt_data() for more details.

def func(x):
    nx = len(x)
    if nx == 0:
        return ""
    elif nx < 10:
        return sorted(x.unique())
    else:
        return f"..."

(
    df.utc_time
    .value_counts()
    .sort_index()
    .to_frame()
    .assign(
        wbans=df.groupby("utc_time")["wban"].apply(func),
    )
)

	count	wbans
utc_time
2026-05-27 20:00:00	1	[53878]
2026-05-27 21:00:00	3	[53878, 63855, 94995]
2026-05-27 22:00:00	4	[53878, 63855, 94995]
2026-05-27 23:00:00	4	[53878, 63855, 94995]
2026-05-28 00:00:00	4	[53878, 63855, 94995]
...	...	...
2026-06-01 20:00:00	152	...
2026-06-01 21:00:00	150	...
2026-06-01 22:00:00	150	...
2026-06-01 23:00:00	150	...
2026-06-02 00:00:00	148	...

125 rows × 2 columns

In these files, for example, site WBAN 13301 has data for multiple earlier days included.

• https://www.ncei.noaa.gov/pub/data/uscrn/products/hourly02/updates/2024/CRN60H0203-202402212000.txt

• https://www.ncei.noaa.gov/pub/data/uscrn/products/hourly02/updates/2024/CRN60H0203-202402222000.txt

Plot temperature change time series

fig, ax = plt.subplots(figsize=(7, 4))

df_ = df.copy()

tmax = df_.utc_time.max()
df_ = df_[df_.utc_time.between(tmax - pd.Timedelta("7h"), tmax)]

df_["utc_time_mid"] = df_["utc_time"] + pd.Timedelta("30min")
df_["t_hr_avg_k"] = df_["t_hr_avg"].add(273.15)
df_["dt_hr_avg"] = df_["t_hr_avg_k"].sub(df_.groupby("wban")["t_hr_avg_k"].transform("mean"))

df_[["utc_time", "dt_hr_avg"]].groupby("utc_time").mean().plot(
    color="0.3",
    linewidth=3,
    zorder=10,
    legend=False,
    ax=ax,
)

df_.groupby("wban").plot(
    x="utc_time",
    y="dt_hr_avg",
    color="0.5",
    linewidth=1,
    alpha=0.4,
    legend=False,
    xlabel="Time (UTC)",
    ylabel="NRT temperature anomaly  $\Delta T$  (°C)",
    ax=ax,
)

ax.set_title(df.attrs["title"], loc="left", size=8);

Plot current temperature

fig, ax = plt.subplots(figsize=(7, 4.5))

ds = uscrn.to_xarray(df)

ds.isel(time=-1).plot.scatter(x="longitude", y="latitude", hue="t_hr_avg", ax=ax);

Specific period of hourly data

Date selection works by file, not by the data inside the file. In general, the data are an hour behind the file date/time. See the notes in uscrn.get_nrt_data() for more details.

def get_nrt_hourly_period(period):
    a, b = period
    ap1 = pd.to_datetime(a) + pd.Timedelta(hours=1)
    bp1 = pd.to_datetime(b) + pd.Timedelta(hours=1)

    df = uscrn.get_nrt_data((ap1, bp1))

    time.sleep(0.5)  # for prints
    in_period = df.utc_time.between(a, b)
    print(
        f"Got {in_period.sum()}/{len(df)} ({in_period.sum() / len(df):.1%}) records "
        f"in desired period {a} to {b}"
    )
    outside = df.loc[~in_period, "utc_time"].value_counts()
    print(
        "Outside counts:",
        ", ".join(f"{time:%Y-%m-%d %H} ({count})" for time, count in outside.items())
    )

    dupe = df.duplicated(["wban", "utc_time"], keep=False)
    print(f"Got {dupe.sum()} ({dupe.sum() / len(df):.1%}) duplicates")

    return (
        df[in_period]
        .drop_duplicates(["wban", "utc_time"], keep="last")
        .reset_index(drop=True)
    )


df = get_nrt_hourly_period(("2024-02-09 16", "2024-02-09 20"))

Show code cell output

Hide code cell output

/home/docs/checkouts/readthedocs.org/user_builds/uscrn/envs/latest/lib/python3.11/site-packages/uscrn/data.py:778: UserWarning: Timestamp 2024-02-09 17:00:00 has no timezone, assuming UTC.
  warnings.warn(f"Timestamp {ts} has no timezone, assuming UTC.")
/home/docs/checkouts/readthedocs.org/user_builds/uscrn/envs/latest/lib/python3.11/site-packages/uscrn/data.py:778: UserWarning: Timestamp 2024-02-09 21:00:00 has no timezone, assuming UTC.
  warnings.warn(f"Timestamp {ts} has no timezone, assuming UTC.")

Discovering files...

  Looking for files in these years
  - 2024

Found 5 file(s) to load
https://www.ncei.noaa.gov/pub/data/uscrn/products/hourly02/updates/2024/CRN60H0203-202402091700.txt
...
https://www.ncei.noaa.gov/pub/data/uscrn/products/hourly02/updates/2024/CRN60H0203-202402092100.txt
Reading files...

[Parallel(n_jobs=2)]: Using backend LokyBackend with 2 concurrent workers.

[Parallel(n_jobs=2)]: Done   1 tasks      | elapsed:    0.3s

[Parallel(n_jobs=2)]: Done   3 out of   5 | elapsed:    0.5s remaining:    0.3s
[Parallel(n_jobs=2)]: Done   5 out of   5 | elapsed:    0.7s finished

Got 769/771 (99.7%) records in desired period 2024-02-09 16 to 2024-02-09 20
Outside counts: 2024-02-09 15 (2)
Got 6 (0.8%) duplicates

df

	wban	utc_time	lst_time	crx_vn	longitude	latitude	t_calc	t_hr_avg	t_max	t_min	...	soil_moisture_5	soil_moisture_10	soil_moisture_20	soil_moisture_50	soil_moisture_100	soil_temp_5	soil_temp_10	soil_temp_20	soil_temp_50	soil_temp_100
0	03047	2024-02-09 16:00:00	2024-02-09 10:00:00	2.622	-102.809998	31.620001	11.300000	10.0	11.300000	8.300000	...	0.048	0.038	0.056	0.045	0.035	9.3	9.1	9.5	12.1	13.3
1	03048	2024-02-09 16:00:00	2024-02-09 09:00:00	2.622	-106.889999	34.360001	3.000000	2.3	3.100000	0.900000	...	0.101	0.140	0.109	0.056	0.057	3.1	4.4	6.2	8.2	9.2
2	03054	2024-02-09 16:00:00	2024-02-09 10:00:00	2.622	-102.769997	33.959999	6.700000	6.5	6.700000	5.900000	...	0.132	0.160	0.164	0.105	0.125	5.0	5.8	6.9	9.0	10.3
3	03055	2024-02-09 16:00:00	2024-02-09 10:00:00	2.622	-101.589996	36.599998	8.700000	6.7	8.700000	3.900000	...	0.250	0.306	0.324	0.412	0.217	4.5	5.0	5.6	6.4	6.6
4	03060	2024-02-09 16:00:00	2024-02-09 09:00:00	2.622	-107.690002	38.540001	-7.400000	-8.4	-7.400000	-10.400000	...	NaN	NaN	NaN	0.173	0.152	-0.3	-0.3	-0.1	0.7	2.0
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
761	94996	2024-02-09 20:00:00	2024-02-09 14:00:00	2.622	-96.849998	40.700001	8.200000	7.8	8.400000	7.000000	...	0.290	0.330	0.287	0.260	0.322	5.1	4.4	4.8	5.2	5.5
762	96404	2024-02-09 20:00:00	2024-02-09 11:00:00	2.514	-141.210007	62.740002	-21.700001	-22.9	-21.700001	-24.200001	...	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN
763	96405	2024-02-09 20:00:00	2024-02-09 11:00:00	2.514	-145.350006	60.470001	3.000000	2.9	3.200000	2.600000	...	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN
764	96407	2024-02-09 20:00:00	2024-02-09 11:00:00	2.515	-159.000000	66.559998	-9.700000	-9.3	-8.700000	-11.100000	...	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN
765	96409	2024-02-09 20:00:00	2024-02-09 11:00:00	2.514	-149.399994	68.650002	-10.800000	-9.6	-8.000000	-11.100000	...	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN

766 rows × 36 columns

Recent daily data

Here, we load the most recent daily data file.

df = uscrn.get_nrt_data(-1, "daily")

Show code cell output

Hide code cell output

Discovering files...

  Looking for files in these years
  - 2026

Found 1 file(s) to load
https://www.ncei.noaa.gov/pub/data/uscrn/products/daily01/updates/2026/CRND0103-202605312359.txt
Reading files...

[Parallel(n_jobs=1)]: Done   1 tasks      | elapsed:    0.2s
[Parallel(n_jobs=1)]: Done   1 out of   1 | elapsed:    0.2s finished

df

	wban	lst_date	crx_vn	longitude	latitude	t_daily_max	t_daily_min	t_daily_mean	t_daily_avg	p_daily_calc	...	soil_moisture_5_daily	soil_moisture_10_daily	soil_moisture_20_daily	soil_moisture_50_daily	soil_moisture_100_daily	soil_temp_5_daily	soil_temp_10_daily	soil_temp_20_daily	soil_temp_50_daily	soil_temp_100_daily
0	03047	2026-05-31	2.622	-102.809998	31.620001	36.299999	20.6	28.500000	26.799999	0.0	...	0.045	0.044	0.052	0.068	0.064	31.6	31.000000	30.400000	30.0	28.299999
1	03048	2026-05-31	2.622	-106.889999	34.360001	33.500000	10.0	21.700001	23.600000	0.0	...	0.043	0.060	0.059	0.059	0.065	30.5	29.200001	27.500000	25.5	24.500000
2	03054	2026-05-31	2.622	-102.769997	33.959999	36.000000	12.9	24.500000	25.500000	0.0	...	0.172	0.191	0.155	0.107	0.133	25.5	24.799999	23.700001	22.9	20.799999
3	03055	2026-05-31	2.622	-101.589996	36.599998	33.099998	12.5	22.799999	24.000000	0.0	...	0.205	0.243	0.197	0.231	0.160	21.9	21.600000	20.600000	19.6	19.299999
4	03060	2026-05-31	2.622	-107.690002	38.540001	19.500000	3.9	11.700000	12.300000	0.0	...	0.079	0.145	0.196	0.259	0.201	18.1	16.799999	15.700000	14.2	11.900000
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
145	96404	2026-05-31	2.514	-141.210007	62.740002	14.800000	-1.4	6.700000	8.400000	3.5	...	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN
146	96405	2026-05-31	2.514	-145.350006	60.470001	11.300000	2.5	6.900000	7.000000	9.1	...	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN
147	96407	2026-05-31	2.515	-159.000000	66.559998	11.200000	2.2	6.700000	6.900000	0.0	...	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN
148	96408	2026-05-31	2.515	-150.869995	63.450001	11.400000	3.7	7.500000	7.200000	1.5	...	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN
149	96409	2026-05-31	2.514	-149.399994	68.650002	7.500000	-3.9	1.800000	1.800000	0.0	...	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN

150 rows × 28 columns

fig, ax = plt.subplots(figsize=(7, 4.5))

ds = uscrn.to_xarray(df).squeeze()

ds.plot.scatter(x="longitude", y="latitude", hue="t_daily_max", ax=ax);