Observations and Machine-Learned Models of Near-Surface Permafrost along the Koyukuk River, Alaska, USA

Schwenk, Jon; Piliouras, Anastasia; Rowland, Joel

doi:10.15485/1922517

Title: Observations and Machine-Learned Models of Near-Surface Permafrost along the Koyukuk River, Alaska, USA

Dataset
Other Related Research

Abstract

This dataset contains GeoTIFs (raster) and GeoPackages (vector) that map observations of near-surface permafrost and not-permafrost from a field campaign conducted near the village of Huslia, AK along the Koyukuk River and its floodplain in July 2018. These data were collected as part of a campaign to understand if and how permafrost impacts riverbank erosion. This problem cannot be assessed without knowing where permafrost exists. Permafrost was observed via frost probing (to a maximum depth of one meter), coring (to a maximum depth of two meters) and bank/bar excavations. An additional boat survey was performed wherein expert (Joel Rowland) judgment assessed the presence or absence of distinctive permafrost features (e.g., overhanging tundra mats, thermoerosional niching, ice wedges, active drainage of ice melt from soils). This dataset also contains the input features and results of two machine learning models (random forest and convolutional neural network) that extrapolate the observations to the full floodplain that may be useful for building, testing, or validating other machine-learned permafrost models. Permafrost data are provided as georasters of the same shape and geovectors (polylines/polygons) and are all projected into EPSG:32605. All data can be visualized with a GIS (QGIS, ArcGIS, etc.).

Authors:

;

Los Alamos National Laboratory; Los Alamos National Laboratory
Pennsylvania State University
Los Alamos National Laboratory

Publication Date:: Sun Jan 01 04:00:00 UTC 2023

Research Org.:: Environmental System Science Data Infrastructure for a Virtual Ecosystem

Sponsoring Org.:: U.S. DOE > Office of Science > Biological and Environmental Research (BER); U.S. DOE > Laboratory Directed Research and Development (LDRD) > Los Alamos National Laboratory

Subject:: 54 ENVIRONMENTAL SCIENCES; EARTH SCIENCE > LAND SURFACE > LAND USE/LAND COVER > LAND USE/LAND COVER CLASSIFICATION > VEGETATION INDEX > NORMALIZED DIFFERENCE VEGETATION INDEX (NDVI); floodplains; machine learning; multispectral image; permafrost; permafrost presence; riverbank erosion; surface water; training data

OSTI Identifier:: 1922517

DOI:: https://doi.org/10.15485/1922517

Citation Formats


                    Schwenk, Jon, Piliouras, Anastasia, and Rowland, Joel. Observations and Machine-Learned Models of Near-Surface Permafrost along the Koyukuk River, Alaska, USA.  United States: N. p., 2023. 
        Web.  doi:10.15485/1922517.

Copy to clipboard


                    Schwenk, Jon, Piliouras, Anastasia, & Rowland, Joel. Observations and Machine-Learned Models of Near-Surface Permafrost along the Koyukuk River, Alaska, USA.  United States.  doi:https://doi.org/10.15485/1922517

Copy to clipboard


                    Schwenk, Jon, Piliouras, Anastasia, and Rowland, Joel. 2023.  
"Observations and Machine-Learned Models of Near-Surface Permafrost along the Koyukuk River, Alaska, USA".  United States.  doi:https://doi.org/10.15485/1922517.  https://www.osti.gov/servlets/purl/1922517. Pub date:Sun Jan 01 04:00:00 UTC 2023

Copy to clipboard


                    
@article{osti_1922517,

  title        = {Observations and Machine-Learned Models of Near-Surface Permafrost along the Koyukuk River, Alaska, USA},

  author       = {Schwenk, Jon and Piliouras, Anastasia and Rowland, Joel},

  abstractNote = {This dataset contains GeoTIFs (raster) and GeoPackages (vector) that map observations of near-surface permafrost and not-permafrost from a field campaign conducted near the village of Huslia, AK along the Koyukuk River and its floodplain in July 2018. These data were collected as part of a campaign to understand if and how permafrost impacts riverbank erosion. This problem cannot be assessed without knowing where permafrost exists. Permafrost was observed via frost probing (to a maximum depth of one meter), coring (to a maximum depth of two meters) and bank/bar excavations. An additional boat survey was performed wherein expert (Joel Rowland) judgment assessed the presence or absence of distinctive permafrost features (e.g., overhanging tundra mats, thermoerosional niching, ice wedges, active drainage of ice melt from soils). This dataset also contains the input features and results of two machine learning models (random forest and convolutional neural network) that extrapolate the observations to the full floodplain that may be useful for building, testing, or validating other machine-learned permafrost models. Permafrost data are provided as georasters of the same shape and geovectors (polylines/polygons) and are all projected into EPSG:32605. All data can be visualized with a GIS (QGIS, ArcGIS, etc.).},

  doi          = {10.15485/1922517},

  journal      = {},

  number       = ,

  volume       = ,

  place        = {United States},

  year         = {Sun Jan 01 04:00:00 UTC 2023},

  month        = {Sun Jan 01 04:00:00 UTC 2023}

}

Copy to clipboard

Dataset:

View Dataset

DOI: https://doi.org/10.15485/1922517

Save / Share:

Export Metadata

Save to My Library

Similar records in DOE Data Explorer and OSTI.GOV collections:

Geomorphic mapping and permafrost occurrence on the Koyukuk River floodplain near Huslia, Alaska

Dataset Douglas, Madison ; Blankenship, Rain ; Chadwick, Austin ; ...

Permafrost occurrence measured in-situ and geomorphic maps drawn by hand using high-resolution satellite imagery for the Koyukuk River floodplain near Huslia, Alaska. The dataset was collected to characterize permafrost extent and rates of formation and degradation due to river processes and climate change. Permafrost occurrence and thickness of the active layer was measured using a 1 or 2 m long permafrost probe or through direct observation (coring, digging trench, exposed ice). Measurements were from June 27 - July 8, 2018 and September 26 - October 1, 2022 and are saved as a csv file. Geomorphic maps delineate floodplain landforms, themore »« less
Riverbank temperatures on the Selawik River, Alaska 2010-2012, and Koyukuk River, Alaska June to July 2018.

Dataset Rowland, Joel ; Crosby, Benjamin ; Schwenk, Jonathan ; ...

The data package includes temperature measurements from riverbanks along the Selawik and Koyukuk Rivers in northwest Alaska. The Selawik River data also includes meteorological data collected locally and the Koyukuk River includes bulk density and ice content data from the riverbanks. The Selawik River data was collected between 2010 and 2013 with meteorological data collected for the entire time period and riverbank temperatures collected in shorter time intervals between 2010 to 2012. All of the Selawik River data were collected a single river bend. The Koyukuk River temperature data was collected from late June to early July 2018 at amore »« less
Machine learning predictions of near-surface permafrost extent at Teller 27, Teller 47, and the Kougarok 64 Hillslope sites on the Seward Peninsula, Alaska: Supporting Data

Dataset Thaler, Evan ; Uhlemann, Sebastian ; Rowland, Joel ; ...

Geophysical surveys were conducted at the NGEE Arctic Teller mile marker 27 site, Teller mile marker 47 site, and Kougarok mile marker 64 site during the summers of 2018, 2019, and 2021. Additional data was collected at Teller mile marker 47 during September 2021 and August 2022. These surveys were used to identify locations of near-surface permafrost during the period of maximum seasonal thaw depth for ground truth data used in machine learning predictions of near-surface permafrost extent at each site. This dataset contains CSV files of ground truth observations of near-permafrost presence or absence for each site, where PFmore »« less
Bulk density, grain size, carbon content and isotopes, and nitrogen content of floodplain sediment along the Koyukuk River, Alaska collected in Summer 2018.

Dataset Douglas, Madison ; Lamb, Michael ; Li, Gen ; ...

The dataset is comprised of grain size, bulk density, carbon and nitrogen content, stable carbon isotopes, and radiocarbon content of soils sampled on the floodplain of the Koyukuk River near Huslia, AK. This dataset was collected to characterize sediment properties and floodplain organic carbon stocks in the discontinuous permafrost floodplain of the Koyukuk River. The data help quantify the flux and storage of sediment and carbon between the Koyukuk River and the surrounding floodplain as the river migrates laterally. The samples were collected in June 27 - July 9, 2018. The soil samples come from a total of 24 locationsmore »« less
Modeling Present and Future Permafrost Distribution at Seward Peninsula, Alaska: Modeling Archive

Dataset Debolskiy, Matvey ; Nicolsky, Dmitry ; Hock, Regine ; ...

This dataset contains results from two GIPL runs for RCP4.5 and RCP8.5 across the Seward Peninsula. The model was run on 500 m by 500 m resolution for the period 1901-2100. Meteo variables include forcing air temperature and precipitation as well as modeled maxim and average snow height. This dataset also includes ground temperature at 10 different depths for every grid cell, active layer depth and talik thickness. Auxillary variables include ecotype map and projected latitude and longitude. Ground temperature borehole data that is used for model calibration and validation of GIPL2 for Seward Peninsula 500 m resolution model runsmore »« less

Similar Records