PACE Level 3

The Plankton, Aerosol, Cloud, ocean Ecosystem (PACE) Level-3 products provide globally gridded derived from Level-2 swath observations. Satellite measurements are spatially and temporally aggregated (daily, 8-day, monthly, seasonal) onto regular latitude–longitude grids at two resolutions (~4 km and 0.1°). These products include ocean color variables such as chlorophyll-a, diffuse attenuation (Kd), and hyperspectral remote sensing reflectance (Rrs), along with derived biogeochemical indicators. For this notebook, we use

• PACE_OCI_L3M_Rrs
• PACE_OCI_L3M_AVW

Steps:

• Create a plan for files to use pc.plan()
• Print the plan to check it print(plan.summary())
• Get matchups pc.matchup(plan)

Note: In a virtual machine in AWS us-west-2, where NASA cloud data is, the point matchups are fast. In Colab, say, your comppute is not in the same data region nor provider (Google versus AWS), and the same matchups might take 10x longer. Thus if you have big matchup tasks, 10s of thousands of points, it is wise to do that in AWS us-west-2.

Prerequisites

# install if needed
!pip install point-collocation --quiet

ERROR: pip's dependency resolver does not currently take into account all the packages that are installed. This behaviour is the source of the following dependency conflicts.
xmip 0.7.2 requires xarrayutils, which is not installed.
xmip 0.7.2 requires xgcm<0.7.0, which is not installed.


# Make sure you are logged in
import earthaccess
earthaccess.login()

<earthaccess.auth.Auth at 0x7f880073fd70>

Read in some points

import pandas as pd
time = "2025-04-09"
lat = 30.0
lon = -89.0

df = pd.DataFrame(
    {
        "lat": [lat],
        "lon": [lon],
        "time": [time],
    }
)
df

	lat	lon	time
0	30.0	-89.0	2025-04-09

Create a plan

%%time
import point_collocation as pc
plan = pc.plan(
    df,
    data_source="earthaccess",
    source_kwargs={
        "short_name": "PACE_OCI_L3M_AOP",
        "granule_name": "*.8D.*.4km.*",
    }
)

CPU times: user 20.9 ms, sys: 0 ns, total: 20.9 ms
Wall time: 484 ms

plan.summary()

Plan: 1 points → 1 unique granule(s)
  Points with 0 matches : 0
  Points with >1 matches: 0
  Time buffer: 0 days 00:00:00

First 1 point(s):
  [0] lat=30.0000, lon=-89.0000, time=2025-04-09 00:00:00: 1 match(es)
    → https://obdaac-tea.earthdatacloud.nasa.gov/ob-cumulus-prod-public/PACE_OCI.20250407_20250414.L3m.8D.AOP.V3_2.4km.nc

Look at variables in that dataset

We will open a granule and inspect.

%%time
plan.open_dataset(0)

open_method: {'xarray_open': 'dataset', 'open_kwargs': {'chunks': {}, 'engine': 'h5netcdf', 'decode_timedelta': False}, 'coords': 'auto', 'set_coords': True, 'dim_renames': None, 'auto_align_phony_dims': None, 'merge': None}
Geolocation auto detected with cf_xarray: ('lon', 'lat') — lon dims=('lon',), lat dims=('lat',)
Points columns used: y='lat', x='lon', time='time'
CPU times: user 438 ms, sys: 157 ms, total: 595 ms
Wall time: 1.36 s

<xarray.Dataset> Size: 26GB
Dimensions:     (lat: 4320, lon: 8640, wavelength: 172)
Coordinates:
  * lat         (lat) float32 17kB 89.98 89.94 89.9 ... -89.9 -89.94 -89.98
  * lon         (lon) float32 35kB -180.0 -179.9 -179.9 ... 179.9 179.9 180.0
  * wavelength  (wavelength) float32 688B 346.0 348.5 350.9 ... 716.8 719.3
Data variables:
    Rrs         (lat, lon, wavelength) float32 26GB dask.array<chunksize=(16, 1024, 8), meta=np.ndarray>
    nflh        (lat, lon) float32 149MB dask.array<chunksize=(16, 1024), meta=np.ndarray>
    avw         (lat, lon) float32 149MB dask.array<chunksize=(16, 1024), meta=np.ndarray>
    aot_865     (lat, lon) float32 149MB dask.array<chunksize=(16, 1024), meta=np.ndarray>
    angstrom    (lat, lon) float32 149MB dask.array<chunksize=(16, 1024), meta=np.ndarray>
Attributes: (12/55)
    product_name:                      PACE_OCI.20250407_20250414.L3m.8D.AOP....
    instrument:                        OCI
    title:                             OCI Level-3 Standard Mapped Image
    project:                           Ocean Biology Processing Group (NASA/G...
    platform:                          PACE
    source:                            satellite observations from OCI-PACE
    ...                                ...
    keywords:                          Earth Science > Oceans > Ocean Optics ...
    id:                                L3/PACE_OCI.20250407_20250414.L3m.8D.A...
    history:                           /sdps/sdpsoper/Science/OCSSW/V2026.2/b...
    processing_version:                3.2
    identifier_product_doi_authority:  https://dx.doi.org
    identifier_product_doi:            10.5067/PACE/OCI/L3M/OC_AOP/3.2

xarray.Dataset

• Dimensions:
  • lat: 4320
  • lon: 8640
  • wavelength: 172
• Coordinates: (3)
  • lat
    (lat)
    float32
    89.98 89.94 89.9 ... -89.94 -89.98

    long_name :

    Latitude

    units :

    degrees_north

    standard_name :

    latitude

    valid_min :

    -90.0

    valid_max :

    90.0

    array([ 89.979164,  89.9375  ,  89.895836, ..., -89.895836, -89.93751 ,
           -89.97917 ], shape=(4320,), dtype=float32)

  • lon
    (lon)
    float32
    -180.0 -179.9 ... 179.9 180.0

    long_name :

    Longitude

    units :

    degrees_east

    standard_name :

    longitude

    valid_min :

    -180.0

    valid_max :

    180.0

    array([-179.97917, -179.9375 , -179.89583, ...,  179.89584,  179.93752,
            179.97917], shape=(8640,), dtype=float32)

  • wavelength
    (wavelength)
    float32
    346.0 348.5 350.9 ... 716.8 719.3

    long_name :

    wavelengths

    units :

    nm

    valid_min :

    0.0

    valid_max :

    20000.0

    array([346.017, 348.468, 350.912, 353.344, 355.782, 358.235, 360.695, 363.137,
           365.61 , 368.083, 370.534, 372.991, 375.482, 377.926, 380.419, 382.876,
           385.359, 387.811, 390.297, 392.764, 395.238, 397.706, 400.178, 402.654,
           405.127, 407.605, 410.074, 412.557, 415.025, 417.512, 419.988, 422.453,
           424.94 , 427.398, 429.885, 432.379, 434.869, 437.351, 439.828, 442.327,
           444.811, 447.309, 449.795, 452.28 , 454.769, 457.262, 459.757, 462.252,
           464.737, 467.244, 469.729, 472.202, 474.7  , 477.189, 479.689, 482.183,
           484.689, 487.182, 489.674, 492.176, 494.686, 497.182, 499.688, 502.19 ,
           504.695, 507.198, 509.72 , 512.213, 514.729, 517.219, 519.747, 522.249,
           524.771, 527.276, 529.798, 532.314, 534.859, 537.346, 539.878, 542.395,
           544.904, 547.441, 549.994, 552.511, 555.044, 557.576, 560.104, 562.642,
           565.19 , 567.71 , 570.259, 572.796, 575.343, 577.902, 580.45 , 582.996,
           585.553, 588.086, 612.73 , 615.145, 617.605, 620.061, 622.53 , 624.988,
           627.434, 629.898, 632.376, 634.83 , 637.305, 639.791, 641.029, 642.255,
           643.479, 644.716, 645.966, 647.188, 648.435, 649.667, 650.913, 652.153,
           653.388, 654.622, 655.869, 657.101, 658.34 , 659.6  , 660.833, 662.067,
           663.3  , 664.564, 665.795, 667.023, 668.263, 669.518, 670.755, 671.99 ,
           673.245, 674.503, 675.731, 676.963, 678.208, 679.448, 680.68 , 681.919,
           683.171, 684.417, 685.657, 686.894, 688.143, 689.394, 690.647, 691.888,
           693.13 , 694.382, 695.644, 696.891, 698.118, 699.376, 700.612, 701.858,
           703.097, 704.354, 705.593, 706.833, 708.089, 709.337, 710.581, 711.826,
           713.068, 714.316, 716.817, 719.298], dtype=float32)

• Data variables: (5)
  • Rrs
    (lat, lon, wavelength)
    float32
    dask.array<chunksize=(16, 1024, 8), meta=np.ndarray>

    long_name :

    Remote sensing reflectance

    units :

    sr^-1

    standard_name :

    surface_ratio_of_upwelling_radiance_emerging_from_sea_water_to_downwelling_radiative_flux_in_air

    valid_min :

    -30000

    valid_max :

    25000

    display_scale :

    linear

    display_min :

    0.0

    display_max :

    0.025

    Array Chunk Bytes 23.92 GiB 512.00 kiB Shape (4320, 8640, 172) (16, 1024, 8) Dask graph 53460 chunks in 2 graph layers Data type float32 numpy.ndarray

  • nflh
    (lat, lon)
    float32
    dask.array<chunksize=(16, 1024), meta=np.ndarray>

    long_name :

    Normalized Fluorescence Line Height

    units :

    W m^-2 um^-1 sr^-1

    valid_min :

    -26470

    valid_max :

    32352

    reference :

    Behrenfeld, M.J., Westberry, T.K., Boss, E.S., O'Malley, R.T., Siegel, D.A., Wiggert, J.D., Franz, B.A., McClain, C.R., Feldman, G.C., Doney, S.C., Moore, J.K., Dall'Olmo, G., Milligan, A.J., Lima, I., and Mahowald, N.: Satellite-detected fluorescence reveals global physiology of ocean phytoplankton, Biogeosciences, 6, 779-794, doi:10.5194/bg-6-779-2009, 2009.

    display_scale :

    linear

    display_min :

    0.0

    display_max :

    0.5

    Array Chunk Bytes 142.38 MiB 64.00 kiB Shape (4320, 8640) (16, 1024) Dask graph 2430 chunks in 2 graph layers Data type float32 numpy.ndarray

  • avw
    (lat, lon)
    float32
    dask.array<chunksize=(16, 1024), meta=np.ndarray>

    long_name :

    Apparent Visible Wavelength

    units :

    nm

    valid_min :

    400.0

    valid_max :

    700.0

    reference :

    Vandermeulen, R. A., Mannino, A., Craig, S.E., Werdell, P.J., 2020: 150 shades of green: Using the full spectrum of remote sensing reflectance to elucidate color shifts in the ocean, Remote Sensing of Environment, 247, 111900, https://doi.org/10.1016/j.rse.2020.111900, https://doi.org/10.5067/KAROCHG01RYJ

    display_scale :

    linear

    display_min :

    450.0

    display_max :

    575.0

    Array Chunk Bytes 142.38 MiB 64.00 kiB Shape (4320, 8640) (16, 1024) Dask graph 2430 chunks in 2 graph layers Data type float32 numpy.ndarray

  • aot_865
    (lat, lon)
    float32
    dask.array<chunksize=(16, 1024), meta=np.ndarray>

    long_name :

    Aerosol optical thickness at 865 nm

    standard_name :

    atmosphere_optical_thickness_due_to_ambient_aerosol

    valid_min :

    0

    valid_max :

    30000

    display_scale :

    linear

    display_min :

    0.0

    display_max :

    0.25

    wavelength :

    864.61

    Array Chunk Bytes 142.38 MiB 64.00 kiB Shape (4320, 8640) (16, 1024) Dask graph 2430 chunks in 2 graph layers Data type float32 numpy.ndarray

  • angstrom
    (lat, lon)
    float32
    dask.array<chunksize=(16, 1024), meta=np.ndarray>

    long_name :

    Aerosol Angstrom exponent, 443 to 865 nm

    standard_name :

    aerosol_angstrom_exponent

    valid_min :

    -30000

    valid_max :

    5000

    display_scale :

    linear

    display_min :

    -0.15

    display_max :

    2.2

    Array Chunk Bytes 142.38 MiB 64.00 kiB Shape (4320, 8640) (16, 1024) Dask graph 2430 chunks in 2 graph layers Data type float32 numpy.ndarray

• Attributes: (55)

  product_name :

  PACE_OCI.20250407_20250414.L3m.8D.AOP.V3_2.4km.nc

  instrument :

  OCI

  title :

  OCI Level-3 Standard Mapped Image

  project :

  Ocean Biology Processing Group (NASA/GSFC/OBPG)

  platform :

  PACE

  source :

  satellite observations from OCI-PACE

  temporal_range :

  7-day

  time_coverage_start :

  2025-04-07T23:41:36.164Z

  time_coverage_end :

  2025-04-15T02:16:01.037Z

  start_orbit_number :

  0

  end_orbit_number :

  0

  map_projection :

  Equidistant Cylindrical

  latitude_units :

  degrees_north

  longitude_units :

  degrees_east

  northernmost_latitude :

  90.0

  southernmost_latitude :

  -90.0

  westernmost_longitude :

  -180.0

  easternmost_longitude :

  180.0

  geospatial_lat_max :

  90.0

  geospatial_lat_min :

  -90.0

  geospatial_lon_max :

  180.0

  geospatial_lon_min :

  -180.0

  latitude_step :

  0.041666668

  longitude_step :

  0.041666668

  sw_point_latitude :

  -89.979164

  sw_point_longitude :

  -179.97917

  spatialResolution :

  4.638312 km

  geospatial_lon_resolution :

  4.638312 km

  geospatial_lat_resolution :

  4.638312 km

  geospatial_lat_units :

  degrees_north

  geospatial_lon_units :

  degrees_east

  number_of_lines :

  4320

  number_of_columns :

  8640

  measure :

  Mean

  Conventions :

  CF-1.6 ACDD-1.3

  institution :

  NASA Goddard Space Flight Center, Ocean Ecology Laboratory, Ocean Biology Processing Group

  standard_name_vocabulary :

  CF Standard Name Table v36

  naming_authority :

  gov.nasa.gsfc.sci.oceandata

  license :

  https://science.nasa.gov/earth-science/earth-science-data/data-information-policy/

  creator_name :

  NASA/GSFC/OBPG

  publisher_name :

  NASA/GSFC/OBPG

  creator_email :

  data@oceancolor.gsfc.nasa.gov

  publisher_email :

  data@oceancolor.gsfc.nasa.gov

  creator_url :

  https://oceandata.sci.gsfc.nasa.gov

  publisher_url :

  https://oceandata.sci.gsfc.nasa.gov

  processing_level :

  L3 Mapped

  cdm_data_type :

  grid

  keywords_vocabulary :

  NASA Global Change Master Directory (GCMD) Science Keywords

  date_created :

  2026-05-16T14:44:01.618904Z

  keywords :

  Earth Science > Oceans > Ocean Optics > Reflectance; Earth Science > Oceans > Ocean Optics > Fluorescence

  id :

  L3/PACE_OCI.20250407_20250414.L3m.8D.AOP.V3_2.4km.nc

  history :

  /sdps/sdpsoper/Science/OCSSW/V2026.2/bin/l3mapmerge -v l3mapmerge_inputs.txt PACE_OCI.20250407_20250414.L3m.8D.AOP.V3_2.4km.nc [2026-05-16T14:47:40] ncattredit.py PACE_OCI.20250407_20250414.L3m.8D.AOP.V3_2.4km.nc /tmp/PACE_OCI.20250407_20250414.L3m.8D.AOP.V3_2.4km.nc.2148470

  processing_version :

  3.2

  identifier_product_doi_authority :

  https://dx.doi.org

  identifier_product_doi :

  10.5067/PACE/OCI/L3M/OC_AOP/3.2

Get the matchups

For variables with a 3rd dimension, like wavelength, all variables will be shown with _3rd dim value. The lat, lon, and time for the matching granules is added as a column. pc_id is the point id/row from the data you are matching. This is added in case there are multiple granules (files) per data point.|

%%time
res = pc.matchup(plan, variables=["Rrs"])
res

CPU times: user 875 ms, sys: 244 ms, total: 1.12 s
Wall time: 2.38 s

	lat	lon	time	pc_id	granule_id	granule_time	granule_lat	granule_lon	Rrs_346	Rrs_348	...	Rrs_705	Rrs_706	Rrs_708	Rrs_709	Rrs_710	Rrs_711	Rrs_713	Rrs_714	Rrs_716	Rrs_719
0	30.0	-89.0	2025-04-09	0	https://obdaac-tea.earthdatacloud.nasa.gov/ob-...	2025-04-10 23:59:59+00:00	30.020832	-89.020828	-0.000272	-0.000106	...	0.003782	0.00368	0.003564	0.00344	0.003306	0.003166	0.003012	0.002802	0.002206	0.001736

1 rows × 180 columns

What if you only want some Rrs wavelengths?

You can filter the dataframe.

res[['lat', 'lon', 'time', 'Rrs_348', 'Rrs_711']]

	lat	lon	time	Rrs_348	Rrs_711
0	30.0	-89.0	2025-04-09	-0.000106	0.003166

Also match on wavelength

point-collocation is designed to match lat/lon/time but you can also match other coordinates that appear in the data. depth, wavelength are common examples. For wavelength, using the filtering above probably makes mose sense, but imagine that you wanted different wavelengths for different locations. To do this, we 2 things:

• The additional coordinate as a column in our dataframe.
• A coord_spec dict that says what the extra coordinate we want to match is.

*Note. In an xarray Dataset, you will see the coordinates in the data variable information, like Rrs (lat, lon, wavelength). Inside the parentheses are the coordinates for that variable.

import pandas as pd
df = pd.DataFrame(
    {
        "lat": [30.0, 31.0],
        "lon": [-89.0, -70.0],
        "time": ["2025-04-09", "2025-04-09"],
        "wave": [400,700]
    }
)
df

	lat	lon	time	wave
0	30.0	-89.0	2025-04-09	400
1	31.0	-70.0	2025-04-09	700

Create our coord_spec

You will need to look at the dataset with plan.open_dataset(0) to see what the coordinates are called in the source.

# Add wavelength as something we can match
coord_spec = {
    "wavelength": {"source": "wavelength", "points": "wave"}
}

Now we make a plan and pass in the coord spec

Rrs in the output is just for the wavelength in the wave column.

import point_collocation as pc
plan = pc.plan(
    df,
    data_source="earthaccess",
    source_kwargs={
        "short_name": "PACE_OCI_L3M_AOP",
        "granule_name": "*.8D.*.4km.*",
    }
)
res = pc.matchup(plan, variables=["Rrs"], coord_spec=coord_spec)
res

	lat	lon	time	wave	pc_id	granule_id	granule_time	granule_lat	granule_lon	Rrs
0	30.0	-89.0	2025-04-09	400	0	https://obdaac-tea.earthdatacloud.nasa.gov/ob-...	2025-04-10 23:59:59+00:00	30.020832	-89.020828	0.001892
1	31.0	-70.0	2025-04-09	700	1	https://obdaac-tea.earthdatacloud.nasa.gov/ob-...	2025-04-10 23:59:59+00:00	31.020832	-70.020828	0.000358

Data variables that are 1D (lat, lon)

In this case, just the variable appears, no _xxx, in the returned dataframe.

%%time
import point_collocation as pc
plan = pc.plan(
    df,
    data_source="earthaccess",
    source_kwargs={
        "short_name": "PACE_OCI_L3M_AOP",
        "granule_name": "*.DAY.*.4km.*",
    }
)
res = pc.matchup(plan, variables=["avw"])
res

CPU times: user 545 ms, sys: 131 ms, total: 676 ms
Wall time: 1.63 s

	lat	lon	time	wave	pc_id	granule_id	granule_time	granule_lat	granule_lon	avw
0	30.0	-89.0	2025-04-09	400	0	https://obdaac-tea.earthdatacloud.nasa.gov/ob-...	2025-04-09 11:59:59+00:00	30.020832	-89.020828	549.833496
1	31.0	-70.0	2025-04-09	700	1	https://obdaac-tea.earthdatacloud.nasa.gov/ob-...	2025-04-09 11:59:59+00:00	31.020832	-70.020828	NaN

Plan with many files

If you are not sure what files to use, you can use a short name without granule_name. Then look at the plan summary to see the file names. You just need to look at one file (n=1). In this example, there are 16 files that match. 2 resolutions (4km and 0.1 deg) and 8 temporal resolutions:

• R32: rolling 32 days starting every 7 days, 4 dates
• SNSP: seasonal/quarterly
• 8D: 8 day
• DAY: daily
• MO: monthly starting 1st day of each month to last

%%time
import point_collocation as pc
plan = pc.plan(
    df,
    data_source="earthaccess",
    source_kwargs={
        "short_name": "PACE_OCI_L3M_AOP",
    }
)

CPU times: user 17.6 ms, sys: 1.63 ms, total: 19.3 ms
Wall time: 410 ms

plan.summary(n=1)

Plan: 2 points → 6 unique granule(s)
  Points with 0 matches : 0
  Points with >1 matches: 2
  Time buffer: 0 days 00:00:00

First 1 point(s):
  [0] lat=30.0000, lon=-89.0000, time=2025-04-09 00:00:00: 6 match(es)
    → https://obdaac-tea.earthdatacloud.nasa.gov/ob-cumulus-prod-public/PACE_OCI.20250401_20250430.L3m.MO.AOP.V3_2.0p1deg.nc
    → https://obdaac-tea.earthdatacloud.nasa.gov/ob-cumulus-prod-public/PACE_OCI.20250401_20250430.L3m.MO.AOP.V3_2.4km.nc
    → https://obdaac-tea.earthdatacloud.nasa.gov/ob-cumulus-prod-public/PACE_OCI.20250407_20250414.L3m.8D.AOP.V3_2.0p1deg.nc
    → https://obdaac-tea.earthdatacloud.nasa.gov/ob-cumulus-prod-public/PACE_OCI.20250407_20250414.L3m.8D.AOP.V3_2.4km.nc
    → https://obdaac-tea.earthdatacloud.nasa.gov/ob-cumulus-prod-public/PACE_OCI.20250409.L3m.DAY.AOP.V3_2.0p1deg.nc
    → https://obdaac-tea.earthdatacloud.nasa.gov/ob-cumulus-prod-public/PACE_OCI.20250409.L3m.DAY.AOP.V3_2.4km.nc

Filter to the files you want

Once you see the files names, you can filter to the ones you want. using granule_name. For example *.SNSP.*.4km.* to get the seasonal (quarterly) values. * are wildcard values.

%%time
import point_collocation as pc
plan = pc.plan(
    df,
    data_source="earthaccess",
    source_kwargs={
        "short_name": "PACE_OCI_L3M_AOP",
        "granule_name": "*.DAY.*.4km.*"
    }
)

CPU times: user 19.1 ms, sys: 258 μs, total: 19.3 ms
Wall time: 440 ms

plan.summary()

Plan: 2 points → 1 unique granule(s)
  Points with 0 matches : 0
  Points with >1 matches: 0
  Time buffer: 0 days 00:00:00

First 2 point(s):
  [0] lat=30.0000, lon=-89.0000, time=2025-04-09 00:00:00: 1 match(es)
    → https://obdaac-tea.earthdatacloud.nasa.gov/ob-cumulus-prod-public/PACE_OCI.20250409.L3m.DAY.AOP.V3_2.4km.nc
  [1] lat=31.0000, lon=-70.0000, time=2025-04-09 00:00:00: 1 match(es)
    → https://obdaac-tea.earthdatacloud.nasa.gov/ob-cumulus-prod-public/PACE_OCI.20250409.L3m.DAY.AOP.V3_2.4km.nc

Try many points

import pandas as pd
url = (
    "https://raw.githubusercontent.com/"
    "fish-pace/point-collocation/main/"
    "examples/fixtures/points.csv"
)
df_points = pd.read_csv(url)
print(len(df_points))

# Let's add on our own pc_id column
df_points = df_points.reset_index(drop=True)
df_points["pc_id"] = df_points.index + 1
df_points["pc_label"] = "pace_" + df_points["pc_id"].astype(str)

df_points.head()

595

	lat	lon	date	pc_id	pc_label
0	27.3835	-82.7375	2024-06-13	1	pace_1
1	27.1190	-82.7125	2024-06-14	2	pace_2
2	26.9435	-82.8170	2024-06-14	3	pace_3
3	26.6875	-82.8065	2024-06-14	4	pace_4
4	26.6675	-82.6455	2024-06-14	5	pace_5

Get a plan for matchups from PACE data

For this example, we will just get a plan for the first 100 points so that it runs quickly.

%%time
import point_collocation as pc
plan = pc.plan(
    df_points[0:100],
    data_source="earthaccess",
    source_kwargs={
        "short_name": "PACE_OCI_L3M_AOP",
        "granule_name": "*.DAY.*.4km.*",
    }
)

CPU times: user 40 ms, sys: 0 ns, total: 40 ms
Wall time: 547 ms

plan.summary(n=0)

Plan: 100 points → 18 unique granule(s)
  Points with 0 matches : 0
  Points with >1 matches: 0
  Time buffer: 0 days 00:00:00

plan.open_dataset(0)

open_method: {'xarray_open': 'dataset', 'open_kwargs': {'chunks': {}, 'engine': 'h5netcdf', 'decode_timedelta': False}, 'coords': 'auto', 'set_coords': True, 'dim_renames': None, 'auto_align_phony_dims': None, 'merge': None}
Geolocation auto detected with cf_xarray: ('lon', 'lat') — lon dims=('lon',), lat dims=('lat',)
Points columns used: y='lat', x='lon', time='date'

<xarray.Dataset> Size: 26GB
Dimensions:     (lat: 4320, lon: 8640, wavelength: 172)
Coordinates:
  * lat         (lat) float32 17kB 89.98 89.94 89.9 ... -89.9 -89.94 -89.98
  * lon         (lon) float32 35kB -180.0 -179.9 -179.9 ... 179.9 179.9 180.0
  * wavelength  (wavelength) float32 688B 346.0 348.5 350.9 ... 716.8 719.3
Data variables:
    Rrs         (lat, lon, wavelength) float32 26GB dask.array<chunksize=(16, 1024, 8), meta=np.ndarray>
    nflh        (lat, lon) float32 149MB dask.array<chunksize=(16, 1024), meta=np.ndarray>
    avw         (lat, lon) float32 149MB dask.array<chunksize=(16, 1024), meta=np.ndarray>
    aot_865     (lat, lon) float32 149MB dask.array<chunksize=(16, 1024), meta=np.ndarray>
    angstrom    (lat, lon) float32 149MB dask.array<chunksize=(16, 1024), meta=np.ndarray>
Attributes: (12/55)
    product_name:                      PACE_OCI.20240613.L3m.DAY.AOP.V3_2.4km.nc
    instrument:                        OCI
    title:                             OCI Level-3 Standard Mapped Image
    project:                           Ocean Biology Processing Group (NASA/G...
    platform:                          PACE
    source:                            satellite observations from OCI-PACE
    ...                                ...
    keywords:                          Earth Science > Oceans > Ocean Optics ...
    id:                                L3/PACE_OCI.20240613.L3m.DAY.AOP.V3_2....
    history:                           /sdps/sdpsoper/Science/OCSSW/V2026.2/b...
    processing_version:                3.2
    identifier_product_doi_authority:  https://dx.doi.org
    identifier_product_doi:            10.5067/PACE/OCI/L3M/OC_AOP/3.2

xarray.Dataset

• Dimensions:
  • lat: 4320
  • lon: 8640
  • wavelength: 172
• Coordinates: (3)
  • lat
    (lat)
    float32
    89.98 89.94 89.9 ... -89.94 -89.98

    long_name :

    Latitude

    units :

    degrees_north

    standard_name :

    latitude

    valid_min :

    -90.0

    valid_max :

    90.0

    array([ 89.979164,  89.9375  ,  89.895836, ..., -89.895836, -89.93751 ,
           -89.97917 ], shape=(4320,), dtype=float32)

  • lon
    (lon)
    float32
    -180.0 -179.9 ... 179.9 180.0

    long_name :

    Longitude

    units :

    degrees_east

    standard_name :

    longitude

    valid_min :

    -180.0

    valid_max :

    180.0

    array([-179.97917, -179.9375 , -179.89583, ...,  179.89584,  179.93752,
            179.97917], shape=(8640,), dtype=float32)

  • wavelength
    (wavelength)
    float32
    346.0 348.5 350.9 ... 716.8 719.3

    long_name :

    wavelengths

    units :

    nm

    valid_min :

    0.0

    valid_max :

    20000.0

    array([346.017, 348.468, 350.912, 353.344, 355.782, 358.235, 360.695, 363.137,
           365.61 , 368.083, 370.534, 372.991, 375.482, 377.926, 380.419, 382.876,
           385.359, 387.811, 390.297, 392.764, 395.238, 397.706, 400.178, 402.654,
           405.127, 407.605, 410.074, 412.557, 415.025, 417.512, 419.988, 422.453,
           424.94 , 427.398, 429.885, 432.379, 434.869, 437.351, 439.828, 442.327,
           444.811, 447.309, 449.795, 452.28 , 454.769, 457.262, 459.757, 462.252,
           464.737, 467.244, 469.729, 472.202, 474.7  , 477.189, 479.689, 482.183,
           484.689, 487.182, 489.674, 492.176, 494.686, 497.182, 499.688, 502.19 ,
           504.695, 507.198, 509.72 , 512.213, 514.729, 517.219, 519.747, 522.249,
           524.771, 527.276, 529.798, 532.314, 534.859, 537.346, 539.878, 542.395,
           544.904, 547.441, 549.994, 552.511, 555.044, 557.576, 560.104, 562.642,
           565.19 , 567.71 , 570.259, 572.796, 575.343, 577.902, 580.45 , 582.996,
           585.553, 588.086, 612.73 , 615.145, 617.605, 620.061, 622.53 , 624.988,
           627.434, 629.898, 632.376, 634.83 , 637.305, 639.791, 641.029, 642.255,
           643.479, 644.716, 645.966, 647.188, 648.435, 649.667, 650.913, 652.153,
           653.388, 654.622, 655.869, 657.101, 658.34 , 659.6  , 660.833, 662.067,
           663.3  , 664.564, 665.795, 667.023, 668.263, 669.518, 670.755, 671.99 ,
           673.245, 674.503, 675.731, 676.963, 678.208, 679.448, 680.68 , 681.919,
           683.171, 684.417, 685.657, 686.894, 688.143, 689.394, 690.647, 691.888,
           693.13 , 694.382, 695.644, 696.891, 698.118, 699.376, 700.612, 701.858,
           703.097, 704.354, 705.593, 706.833, 708.089, 709.337, 710.581, 711.826,
           713.068, 714.316, 716.817, 719.298], dtype=float32)

• Data variables: (5)
  • Rrs
    (lat, lon, wavelength)
    float32
    dask.array<chunksize=(16, 1024, 8), meta=np.ndarray>

    long_name :

    Remote sensing reflectance

    units :

    sr^-1

    standard_name :

    surface_ratio_of_upwelling_radiance_emerging_from_sea_water_to_downwelling_radiative_flux_in_air

    valid_min :

    -30000

    valid_max :

    25000

    display_scale :

    linear

    display_min :

    0.0

    display_max :

    0.025

    Array Chunk Bytes 23.92 GiB 512.00 kiB Shape (4320, 8640, 172) (16, 1024, 8) Dask graph 53460 chunks in 2 graph layers Data type float32 numpy.ndarray

  • nflh
    (lat, lon)
    float32
    dask.array<chunksize=(16, 1024), meta=np.ndarray>

    long_name :

    Normalized Fluorescence Line Height

    units :

    W m^-2 um^-1 sr^-1

    valid_min :

    -26470

    valid_max :

    32352

    reference :

    Behrenfeld, M.J., Westberry, T.K., Boss, E.S., O'Malley, R.T., Siegel, D.A., Wiggert, J.D., Franz, B.A., McClain, C.R., Feldman, G.C., Doney, S.C., Moore, J.K., Dall'Olmo, G., Milligan, A.J., Lima, I., and Mahowald, N.: Satellite-detected fluorescence reveals global physiology of ocean phytoplankton, Biogeosciences, 6, 779-794, doi:10.5194/bg-6-779-2009, 2009.

    display_scale :

    linear

    display_min :

    0.0

    display_max :

    0.5

    Array Chunk Bytes 142.38 MiB 64.00 kiB Shape (4320, 8640) (16, 1024) Dask graph 2430 chunks in 2 graph layers Data type float32 numpy.ndarray

  • avw
    (lat, lon)
    float32
    dask.array<chunksize=(16, 1024), meta=np.ndarray>

    long_name :

    Apparent Visible Wavelength

    units :

    nm

    valid_min :

    400.0

    valid_max :

    700.0

    reference :

    Vandermeulen, R. A., Mannino, A., Craig, S.E., Werdell, P.J., 2020: 150 shades of green: Using the full spectrum of remote sensing reflectance to elucidate color shifts in the ocean, Remote Sensing of Environment, 247, 111900, https://doi.org/10.1016/j.rse.2020.111900, https://doi.org/10.5067/KAROCHG01RYJ

    display_scale :

    linear

    display_min :

    450.0

    display_max :

    575.0

    Array Chunk Bytes 142.38 MiB 64.00 kiB Shape (4320, 8640) (16, 1024) Dask graph 2430 chunks in 2 graph layers Data type float32 numpy.ndarray

  • aot_865
    (lat, lon)
    float32
    dask.array<chunksize=(16, 1024), meta=np.ndarray>

    long_name :

    Aerosol optical thickness at 865 nm

    standard_name :

    atmosphere_optical_thickness_due_to_ambient_aerosol

    valid_min :

    0

    valid_max :

    30000

    display_scale :

    linear

    display_min :

    0.0

    display_max :

    0.25

    wavelength :

    864.61

    Array Chunk Bytes 142.38 MiB 64.00 kiB Shape (4320, 8640) (16, 1024) Dask graph 2430 chunks in 2 graph layers Data type float32 numpy.ndarray

  • angstrom
    (lat, lon)
    float32
    dask.array<chunksize=(16, 1024), meta=np.ndarray>

    long_name :

    Aerosol Angstrom exponent, 443 to 865 nm

    standard_name :

    aerosol_angstrom_exponent

    valid_min :

    -30000

    valid_max :

    5000

    display_scale :

    linear

    display_min :

    -0.15

    display_max :

    2.2

    Array Chunk Bytes 142.38 MiB 64.00 kiB Shape (4320, 8640) (16, 1024) Dask graph 2430 chunks in 2 graph layers Data type float32 numpy.ndarray

• Attributes: (55)

  product_name :

  PACE_OCI.20240613.L3m.DAY.AOP.V3_2.4km.nc

  instrument :

  OCI

  title :

  OCI Level-3 Standard Mapped Image

  project :

  Ocean Biology Processing Group (NASA/GSFC/OBPG)

  platform :

  PACE

  source :

  satellite observations from OCI-PACE

  temporal_range :

  day

  time_coverage_start :

  2024-06-13T00:38:10.218Z

  time_coverage_end :

  2024-06-14T01:27:55.032Z

  start_orbit_number :

  0

  end_orbit_number :

  0

  map_projection :

  Equidistant Cylindrical

  latitude_units :

  degrees_north

  longitude_units :

  degrees_east

  northernmost_latitude :

  90.0

  southernmost_latitude :

  -90.0

  westernmost_longitude :

  -180.0

  easternmost_longitude :

  180.0

  geospatial_lat_max :

  90.0

  geospatial_lat_min :

  -90.0

  geospatial_lon_max :

  180.0

  geospatial_lon_min :

  -180.0

  latitude_step :

  0.041666668

  longitude_step :

  0.041666668

  sw_point_latitude :

  -89.979164

  sw_point_longitude :

  -179.97917

  spatialResolution :

  4.638312 km

  geospatial_lon_resolution :

  4.638312 km

  geospatial_lat_resolution :

  4.638312 km

  geospatial_lat_units :

  degrees_north

  geospatial_lon_units :

  degrees_east

  number_of_lines :

  4320

  number_of_columns :

  8640

  measure :

  Mean

  Conventions :

  CF-1.6 ACDD-1.3

  institution :

  NASA Goddard Space Flight Center, Ocean Ecology Laboratory, Ocean Biology Processing Group

  standard_name_vocabulary :

  CF Standard Name Table v36

  naming_authority :

  gov.nasa.gsfc.sci.oceandata

  license :

  https://science.nasa.gov/earth-science/earth-science-data/data-information-policy/

  creator_name :

  NASA/GSFC/OBPG

  publisher_name :

  NASA/GSFC/OBPG

  creator_email :

  data@oceancolor.gsfc.nasa.gov

  publisher_email :

  data@oceancolor.gsfc.nasa.gov

  creator_url :

  https://oceandata.sci.gsfc.nasa.gov

  publisher_url :

  https://oceandata.sci.gsfc.nasa.gov

  processing_level :

  L3 Mapped

  cdm_data_type :

  grid

  keywords_vocabulary :

  NASA Global Change Master Directory (GCMD) Science Keywords

  date_created :

  2026-05-16T14:43:23.612720Z

  keywords :

  Earth Science > Oceans > Ocean Optics > Fluorescence; Earth Science > Oceans > Ocean Optics > Reflectance

  id :

  L3/PACE_OCI.20240613.L3m.DAY.AOP.V3_2.4km.nc

  history :

  /sdps/sdpsoper/Science/OCSSW/V2026.2/bin/l3mapmerge -v l3mapmerge_inputs.txt PACE_OCI.20240613.L3m.DAY.AOP.V3_2.4km.nc [2026-05-16T14:47:19] ncattredit.py PACE_OCI.20240613.L3m.DAY.AOP.V3_2.4km.nc /tmp/PACE_OCI.20240613.L3m.DAY.AOP.V3_2.4km.nc.525114

  processing_version :

  3.2

  identifier_product_doi_authority :

  https://dx.doi.org

  identifier_product_doi :

  10.5067/PACE/OCI/L3M/OC_AOP/3.2

Get 100 matchups using that plan

In a virtual machine in AWS us-west-2, where NASA cloud data is, this is 12 seconds. In Colab, say, this might be over a minute since you are not in the same data region nor provider (Google versus AWS).

%%time
res = pc.matchup(plan, variables = ["avw"])

CPU times: user 6.63 s, sys: 1.15 s, total: 7.78 s
Wall time: 21.7 s

res.head()

	lat	lon	time	pc_id	pc_label	granule_id	granule_time	granule_lat	granule_lon	avw
0	27.3835	-82.7375	2024-06-13 12:00:00	1	pace_1	https://obdaac-tea.earthdatacloud.nasa.gov/ob-...	2024-06-13 11:59:59+00:00	27.395832	-82.729164	NaN
1	27.1190	-82.7125	2024-06-14 12:00:00	2	pace_2	https://obdaac-tea.earthdatacloud.nasa.gov/ob-...	2024-06-14 11:59:59+00:00	27.104164	-82.729164	NaN
2	26.9435	-82.8170	2024-06-14 12:00:00	3	pace_3	https://obdaac-tea.earthdatacloud.nasa.gov/ob-...	2024-06-14 11:59:59+00:00	26.937498	-82.812500	NaN
3	26.6875	-82.8065	2024-06-14 12:00:00	4	pace_4	https://obdaac-tea.earthdatacloud.nasa.gov/ob-...	2024-06-14 11:59:59+00:00	26.687498	-82.812500	NaN
4	26.6675	-82.6455	2024-06-14 12:00:00	5	pace_5	https://obdaac-tea.earthdatacloud.nasa.gov/ob-...	2024-06-14 11:59:59+00:00	26.687498	-82.645828	NaN

Try lots of products

Pick a recent data point so NRT works. Not all products have files.

import pandas as pd
time = "2026-01-09"
lat = 30.0
lon = -89.0

df = pd.DataFrame(
    {
        "lat": [lat],
        "lon": [lon],
        "time": [time],
    }
)
df["time"] = pd.to_datetime(df["time"])

import earthaccess
results = earthaccess.search_datasets(instrument="oci")

short_names = [
    item.summary()["short-name"]
    for item in results
    if "L3M" in item.summary()["short-name"]
]

print(short_names)

['PACE_OCI_L3M_UVAI_UAA_NRT', 'PACE_OCI_L3M_UVAI_UAA', 'PACE_OCI_L3M_AER_UAA_NRT', 'PACE_OCI_L3M_AER_UAA', 'PACE_OCI_L3M_AOP_NRT', 'PACE_OCI_L3M_AOP', 'PACE_OCI_L3M_CLOSE', 'PACE_OCI_L3M_CLOUD_MASK_NRT', 'PACE_OCI_L3M_CLOUD_MASK', 'PACE_OCI_L3M_CLOUD_NRT', 'PACE_OCI_L3M_CLOUD', 'PACE_OCI_L3M_KD_NRT', 'PACE_OCI_L3M_KD', 'PACE_OCI_L3M_LANDVI_NRT', 'PACE_OCI_L3M_LANDVI', 'PACE_OCI_L3M_BGC_NRT', 'PACE_OCI_L3M_BGC', 'PACE_OCI_L3M_IOP_NRT', 'PACE_OCI_L3M_IOP', 'PACE_OCI_L3M_PAR_NRT', 'PACE_OCI_L3M_PAR', 'PACE_OCI_L3M_SFREFL_NRT', 'PACE_OCI_L3M_SFREFL', 'PACE_OCI_L3M_TRGAS_NRT', 'PACE_OCI_L3M_TRGAS']

%%time
# Confirm works for all L3 products
# Good. PACE_OCI_L3M_TRGAS is slow.
import point_collocation as pc
for short_name in short_names:
    print(f"\n===== {short_name} =====")

    try:
        plan = pc.plan(
            df,
            data_source="earthaccess",
            source_kwargs={
                "short_name": short_name,
                "granule_name":"*.DAY.*",
             }
        )
        plan.open_dataset(0)
    except Exception as e:
        print("Failed:", e)

===== PACE_OCI_L3M_UVAI_UAA_NRT =====
Failed: result index 0 is out of range for a plan with 0 result(s). Valid indices are 0 to -1.

===== PACE_OCI_L3M_UVAI_UAA =====
Failed: result index 0 is out of range for a plan with 0 result(s). Valid indices are 0 to -1.

===== PACE_OCI_L3M_AER_UAA_NRT =====
open_method: {'xarray_open': 'dataset', 'open_kwargs': {'chunks': {}, 'engine': 'h5netcdf', 'decode_timedelta': False}, 'coords': 'auto', 'set_coords': True, 'dim_renames': None, 'auto_align_phony_dims': None, 'merge': None}
Geolocation auto detected with cf_xarray: ('lon', 'lat') — lon dims=('lon',), lat dims=('lat',)
Points columns used: y='lat', x='lon', time='time'

===== PACE_OCI_L3M_AER_UAA =====
open_method: {'xarray_open': 'dataset', 'open_kwargs': {'chunks': {}, 'engine': 'h5netcdf', 'decode_timedelta': False}, 'coords': 'auto', 'set_coords': True, 'dim_renames': None, 'auto_align_phony_dims': None, 'merge': None}
Geolocation auto detected with cf_xarray: ('lon', 'lat') — lon dims=('lon',), lat dims=('lat',)
Points columns used: y='lat', x='lon', time='time'

===== PACE_OCI_L3M_AOP_NRT =====
Failed: result index 0 is out of range for a plan with 0 result(s). Valid indices are 0 to -1.

===== PACE_OCI_L3M_AOP =====
open_method: {'xarray_open': 'dataset', 'open_kwargs': {'chunks': {}, 'engine': 'h5netcdf', 'decode_timedelta': False}, 'coords': 'auto', 'set_coords': True, 'dim_renames': None, 'auto_align_phony_dims': None, 'merge': None}
Geolocation auto detected with cf_xarray: ('lon', 'lat') — lon dims=('lon',), lat dims=('lat',)
Points columns used: y='lat', x='lon', time='time'

===== PACE_OCI_L3M_CLOSE =====
Failed: result index 0 is out of range for a plan with 0 result(s). Valid indices are 0 to -1.

===== PACE_OCI_L3M_CLOUD_MASK_NRT =====
Failed: result index 0 is out of range for a plan with 0 result(s). Valid indices are 0 to -1.

===== PACE_OCI_L3M_CLOUD_MASK =====
Failed: result index 0 is out of range for a plan with 0 result(s). Valid indices are 0 to -1.

===== PACE_OCI_L3M_CLOUD_NRT =====
open_method: {'xarray_open': 'dataset', 'open_kwargs': {'chunks': {}, 'engine': 'h5netcdf', 'decode_timedelta': False}, 'coords': 'auto', 'set_coords': True, 'dim_renames': None, 'auto_align_phony_dims': None, 'merge': None}
Geolocation auto detected with cf_xarray: ('lon', 'lat') — lon dims=('lon',), lat dims=('lat',)
Points columns used: y='lat', x='lon', time='time'

===== PACE_OCI_L3M_CLOUD =====
open_method: {'xarray_open': 'dataset', 'open_kwargs': {'chunks': {}, 'engine': 'h5netcdf', 'decode_timedelta': False}, 'coords': 'auto', 'set_coords': True, 'dim_renames': None, 'auto_align_phony_dims': None, 'merge': None}
Geolocation auto detected with cf_xarray: ('lon', 'lat') — lon dims=('lon',), lat dims=('lat',)
Points columns used: y='lat', x='lon', time='time'

===== PACE_OCI_L3M_KD_NRT =====
Failed: result index 0 is out of range for a plan with 0 result(s). Valid indices are 0 to -1.

===== PACE_OCI_L3M_KD =====
open_method: {'xarray_open': 'dataset', 'open_kwargs': {'chunks': {}, 'engine': 'h5netcdf', 'decode_timedelta': False}, 'coords': 'auto', 'set_coords': True, 'dim_renames': None, 'auto_align_phony_dims': None, 'merge': None}
Geolocation auto detected with cf_xarray: ('lon', 'lat') — lon dims=('lon',), lat dims=('lat',)
Points columns used: y='lat', x='lon', time='time'

===== PACE_OCI_L3M_LANDVI_NRT =====
open_method: {'xarray_open': 'dataset', 'open_kwargs': {'chunks': {}, 'engine': 'h5netcdf', 'decode_timedelta': False}, 'coords': 'auto', 'set_coords': True, 'dim_renames': None, 'auto_align_phony_dims': None, 'merge': None}
Geolocation auto detected with cf_xarray: ('lon', 'lat') — lon dims=('lon',), lat dims=('lat',)
Points columns used: y='lat', x='lon', time='time'

===== PACE_OCI_L3M_LANDVI =====
open_method: {'xarray_open': 'dataset', 'open_kwargs': {'chunks': {}, 'engine': 'h5netcdf', 'decode_timedelta': False}, 'coords': 'auto', 'set_coords': True, 'dim_renames': None, 'auto_align_phony_dims': None, 'merge': None}
Geolocation auto detected with cf_xarray: ('lon', 'lat') — lon dims=('lon',), lat dims=('lat',)
Points columns used: y='lat', x='lon', time='time'

===== PACE_OCI_L3M_BGC_NRT =====
Failed: result index 0 is out of range for a plan with 0 result(s). Valid indices are 0 to -1.

===== PACE_OCI_L3M_BGC =====
open_method: {'xarray_open': 'dataset', 'open_kwargs': {'chunks': {}, 'engine': 'h5netcdf', 'decode_timedelta': False}, 'coords': 'auto', 'set_coords': True, 'dim_renames': None, 'auto_align_phony_dims': None, 'merge': None}
Geolocation auto detected with cf_xarray: ('lon', 'lat') — lon dims=('lon',), lat dims=('lat',)
Points columns used: y='lat', x='lon', time='time'

===== PACE_OCI_L3M_IOP_NRT =====
Failed: result index 0 is out of range for a plan with 0 result(s). Valid indices are 0 to -1.

===== PACE_OCI_L3M_IOP =====
open_method: {'xarray_open': 'dataset', 'open_kwargs': {'chunks': {}, 'engine': 'h5netcdf', 'decode_timedelta': False}, 'coords': 'auto', 'set_coords': True, 'dim_renames': None, 'auto_align_phony_dims': None, 'merge': None}
Geolocation auto detected with cf_xarray: ('lon', 'lat') — lon dims=('lon',), lat dims=('lat',)
Points columns used: y='lat', x='lon', time='time'

===== PACE_OCI_L3M_PAR_NRT =====
Failed: result index 0 is out of range for a plan with 0 result(s). Valid indices are 0 to -1.

===== PACE_OCI_L3M_PAR =====
open_method: {'xarray_open': 'dataset', 'open_kwargs': {'chunks': {}, 'engine': 'h5netcdf', 'decode_timedelta': False}, 'coords': 'auto', 'set_coords': True, 'dim_renames': None, 'auto_align_phony_dims': None, 'merge': None}
Geolocation auto detected with cf_xarray: ('lon', 'lat') — lon dims=('lon',), lat dims=('lat',)
Points columns used: y='lat', x='lon', time='time'

===== PACE_OCI_L3M_SFREFL_NRT =====