soilspecdata 0.0.1

Download and load soil spectral data

SoilSpecData

A Python package for handling soil spectroscopy data, with a focus on the Open Soil Spectral Library (OSSL).

Installation

pip install -U soilspecdata

The -U flag is used to update the package to the latest version. This is important to ensure that you have the latest features and bug fixes.

If you want to install the development version, run in the project root:

pip install -e .[dev]

Features

• Easy loading and handling of OSSL dataset
• Support for both VISNIR (Visible Near-Infrared) and MIR (Mid-Infrared) spectral data
• Flexible wavenumber range filtering
• Convenient access to soil properties and metadata
• Automatic caching of downloaded data
• Get aligned spectra and target variable(s)
• Further datasets to come …

Quick Start

# Import the package
from soilspecdata.datasets.ossl import get_ossl

Load OSSL dataset

ossl = get_ossl()

The spectral analysis covers both MIR (400-4000 cm⁻¹) and VISNIR (4000-28571 cm⁻¹) regions, with data reported in increasing wavenumbers for consistency across the entire spectral range.

Ranges of interest can further be filtered using the wmin and wmax parameters in the get_mir and get_visnir methods.

MIR spectra

mir_data = ossl.get_mir()

VISNIR spectra

Using custom wavenumber range:

visnir_data = ossl.get_visnir(wmin=4000, wmax=25000)

VISNIR | MIR dataclass member variables

print(visnir_data)

SpectraData attributes:
----------------------
Available attributes: wavenumbers, spectra, measurement_type, sample_ids

Wavenumbers:
-----------
[4000, 4003, 4006, 4009, 4012, 4016, 4019, 4022, 4025, 4029]
Shape: (1051,)

Spectra:
-------
[[0.3859, 0.3819, 0.3792, 0.3776, 0.3769],
 [0.3429, 0.3419, 0.3414, 0.3413, 0.3415],
 [0.3425, 0.3384, 0.3354, 0.3334, 0.3323],
 [0.2745, 0.2754, 0.2759, 0.2761, 0.276 ],
 [0.285 , 0.2794, 0.2755, 0.273 , 0.2718]]
Shape: (64644, 1051)

Measurement type (Reflectance or Absorbance):
--------------------------------------------
ref

Sample IDs:
----------
['FS15R_FS4068', 'FS15R_FS4069', 'FS15R_FS4070', 'FS15R_FS4071',
 'FS15R_FS4072', 'FS15R_FS4073', 'FS15R_FS4074', 'FS15R_FS4075',
 'FS15R_FS4076', 'FS15R_FS4077']
Total samples: 64644

Getting soil properties and other metadata

Example: get Cation Exchange Capacity (CEC) measurements (in cmolc/kg) for all samples. Results are returned as a pd.DataFrame indexed by sample ID (id):

properties = ossl.get_properties(['cec_usda.a723_cmolc.kg'], require_complete=True)

properties.head()

<style scoped> .dataframe tbody tr th:only-of-type { vertical-align: middle; } .dataframe tbody tr th { vertical-align: top; } .dataframe thead th { text-align: right; } </style>

	cec_usda.a723_cmolc.kg
id
S40857	6.633217
S40858	3.822628
S40859	3.427324
S40860	1.906545
S40861	13.403203

[!NOTE]

require_complete=True ensures that only non null values are returned in selected columns (here cec_usda.a723_cmolc.kg).

For more details on the OSSL dataset and its variables, see the OSSL documentation. Any column name part of the ossl.properties_cols list can be used as a target or metadata variable.

ossl.properties_cols

['dataset.code_ascii_txt',
 'id.layer_local_c',
 'id.layer_uuid_txt',
 'id.project_ascii_txt',
 'id.location_olc_txt',
 'id.dataset.site_ascii_txt',
 'id.scan_local_c',
 'longitude.point_wgs84_dd',
 'latitude.point_wgs84_dd',
 'layer.sequence_usda_uint16',
 'layer.upper.depth_usda_cm',
 'layer.lower.depth_usda_cm',
 'observation.date.begin_iso.8601_yyyy.mm.dd',
 'observation.date.end_iso.8601_yyyy.mm.dd',
 'surveyor.title_utf8_txt',
 'layer.texture_usda_txt',
 'pedon.taxa_usda_txt',
 'horizon.designation_usda_txt',
 'longitude.county_wgs84_dd',
 'latitude.county_wgs84_dd',
 'location.point.error_any_m',
 'location.country_iso.3166_txt',
 'observation.ogc.schema.title_ogc_txt',
 'observation.ogc.schema_idn_url',
 'surveyor.contact_ietf_email',
 'surveyor.address_utf8_txt',
 'dataset.title_utf8_txt',
 'dataset.owner_utf8_txt',
 'dataset.address_idn_url',
 'dataset.doi_idf_url',
 'dataset.license.title_ascii_txt',
 'dataset.license.address_idn_url',
 'dataset.contact.name_utf8_txt',
 'dataset.contact_ietf_email',
 'acidity_usda.a795_cmolc.kg',
 'aggstb_usda.a1_w.pct',
 'al.dith_usda.a65_w.pct',
 'al.ext_aquaregia_g.kg',
 'al.ext_usda.a1056_mg.kg',
 'al.ext_usda.a69_cmolc.kg',
 'al.ox_usda.a59_w.pct',
 'awc.33.1500kPa_usda.c80_w.frac',
 'b.ext_mel3_mg.kg',
 'bd_iso.11272_g.cm3',
 'bd_usda.a21_g.cm3',
 'bd_usda.a4_g.cm3',
 'c.tot_iso.10694_w.pct',
 'c.tot_usda.a622_w.pct',
 'ca.ext_aquaregia_mg.kg',
 'ca.ext_usda.a1059_mg.kg',
 'ca.ext_usda.a722_cmolc.kg',
 'caco3_iso.10693_w.pct',
 'caco3_usda.a54_w.pct',
 'cec_iso.11260_cmolc.kg',
 'cec_usda.a723_cmolc.kg',
 'cf_iso.11464_w.pct',
 'cf_usda.c236_w.pct',
 'clay.tot_iso.11277_w.pct',
 'clay.tot_usda.a334_w.pct',
 'cu.ext_usda.a1063_mg.kg',
 'ec_iso.11265_ds.m',
 'ec_usda.a364_ds.m',
 'efferv_usda.a479_class',
 'fe.dith_usda.a66_w.pct',
 'fe.ext_aquaregia_g.kg',
 'fe.ext_usda.a1064_mg.kg',
 'fe.ox_usda.a60_w.pct',
 'file_sequence',
 'k.ext_aquaregia_mg.kg',
 'k.ext_usda.a1065_mg.kg',
 'k.ext_usda.a725_cmolc.kg',
 'mg.ext_aquaregia_mg.kg',
 'mg.ext_usda.a1066_mg.kg',
 'mg.ext_usda.a724_cmolc.kg',
 'mn.ext_aquaregia_mg.kg',
 'mn.ext_usda.a1067_mg.kg',
 'mn.ext_usda.a70_mg.kg',
 'n.tot_iso.11261_w.pct',
 'n.tot_iso.13878_w.pct',
 'n.tot_usda.a623_w.pct',
 'na.ext_aquaregia_mg.kg',
 'na.ext_usda.a1068_mg.kg',
 'na.ext_usda.a726_cmolc.kg',
 'oc_iso.10694_w.pct',
 'oc_usda.c1059_w.pct',
 'oc_usda.c729_w.pct',
 'p.ext_aquaregia_mg.kg',
 'p.ext_iso.11263_mg.kg',
 'p.ext_usda.a1070_mg.kg',
 'p.ext_usda.a270_mg.kg',
 'p.ext_usda.a274_mg.kg',
 'p.ext_usda.a652_mg.kg',
 'ph.cacl2_iso.10390_index',
 'ph.cacl2_usda.a477_index',
 'ph.cacl2_usda.a481_index',
 'ph.h2o_iso.10390_index',
 'ph.h2o_usda.a268_index',
 's.ext_mel3_mg.kg',
 's.tot_usda.a624_w.pct',
 'sand.tot_iso.11277_w.pct',
 'sand.tot_usda.c405_w.pct',
 'sand.tot_usda.c60_w.pct',
 'silt.tot_iso.11277_w.pct',
 'silt.tot_usda.c407_w.pct',
 'silt.tot_usda.c62_w.pct',
 'wr.10kPa_usda.a414_w.pct',
 'wr.10kPa_usda.a8_w.pct',
 'wr.1500kPa_usda.a417_w.pct',
 'wr.33kPa_usda.a415_w.pct',
 'wr.33kPa_usda.a9_w.pct',
 'zn.ext_usda.a1073_mg.kg',
 'scan.mir.date.begin_iso.8601_yyyy.mm.dd',
 'scan.mir.date.end_iso.8601_yyyy.mm.dd',
 'scan.mir.model.name_utf8_txt',
 'scan.mir.model.code_any_txt',
 'scan.mir.method.optics_any_txt',
 'scan.mir.method.preparation_any_txt',
 'scan.mir.license.title_ascii_txt',
 'scan.mir.license.address_idn_url',
 'scan.mir.doi_idf_url',
 'scan.mir.contact.name_utf8_txt',
 'scan.mir.contact.email_ietf_txt',
 'scan.visnir.date.begin_iso.8601_yyyy.mm.dd',
 'scan.visnir.date.end_iso.8601_yyyy.mm.dd',
 'scan.visnir.model.name_utf8_txt',
 'scan.visnir.model.code_any_txt',
 'scan.visnir.method.optics_any_txt',
 'scan.visnir.method.preparation_any_txt',
 'scan.visnir.license.title_ascii_txt',
 'scan.visnir.license.address_idn_url',
 'scan.visnir.doi_idf_url',
 'scan.visnir.contact.name_utf8_txt',
 'scan.visnir.contact.email_ietf_txt']

• Get metadata (e.g., geographical coordinates):

metadata = ossl.get_properties(['longitude.point_wgs84_dd', 'latitude.point_wgs84_dd'], require_complete=False)

Preparing data for machine learning pipeline

To get directly aligned spectra and target variable(s):

X, y, ids = ossl.get_aligned_data(
    spectra_data=mir_data,
    target_cols='cec_usda.a723_cmolc.kg'
)

X.shape, y.shape, ids.shape

((57064, 1701), (57064, 1), (57064,))

And plot the first 20 MIR spectra:

from matplotlib import pyplot as plt

plt.figure(figsize=(12, 3))
plt.plot(mir_data.wavenumbers, mir_data.spectra[:20,:].T, alpha=0.3, color='steelblue', lw=1)
plt.gca().invert_xaxis()
plt.grid(True, linestyle='--', alpha=0.7)

plt.xlabel('Wavenumber (cm⁻¹)')
plt.ylabel('Absorbance');

Data Structure

The package returns spectra data in a structured format containing:

• Wavenumbers
• Spectra measurements
• Measurement type (reflectance/absorbance)
• Sample IDs

Properties and metadata are returned as pandas DataFrames indexed by sample ID.

Cache Management

By default, the OSSL dataset is cached in ~/.soilspecdata/. To force a fresh download:

ossl = get_ossl(force_download=True)

Contributing

Contributions are welcome! Please feel free to submit a Pull Request.

License

Apache2

Citation(s)

• OSSL Library: Safanelli, J.L., Hengl, T., Parente, L.L., Minarik, R., Bloom, D.E., Todd-Brown, K., Gholizadeh, A., Mendes, W. de S., Sanderman, J., 2025. Open Soil Spectral Library (OSSL): Building reproducible soil calibration models through open development and community engagement. PLOS ONE 20, e0296545. https://doi.org/10.1371/journal.pone.0296545