datasets

This repo contains constructed data sets from the SerpRateAI project.

Contributing a dataset

In order to contribute a dataset, you will need to:

1. Update the README file to include the an entry in the dataset table as well as a subsection with the Data set name describing the dataset itself (e.g., column descriptions)
2. A single data file for the data set, if there are multiple data files these should each be considered their own data set

These should be committed on their own branch and a pull request is made per data set to merge with main.

Data

Data set	time period	Sample rate	Number of Samples	source
Bubbles	May 2019 - Feb 2020	events	2434	Bubble detection catalog from BA1B for Aiken et al., 2022
Daily Precipitation	Jan 2019 - Feb 2020	daily	425	https://code.earthengine.google.com/65cfcd01ee34290615a7c854a00b76f4
Geology	N/A	N/A	690	constructed from AI paper, Aiken et al.
Hourly Precipitation	Jan 01 2017 - Dec 31 2020	hourly	35064	The hourly_precip file in https://code.earthengine.google.com/?accept_repo=users/cassandralem/tigertail-data and Generate_Hourly_Precip.py in dataset_management folder in this repo
Hourly Temperature 2m	Jan 01 2017 - Dec 31 2020	hourly	35064	The hourly_temperature_2m file in https://code.earthengine.google.com/?accept_repo=users/cassandralem/tigertail-data and the Generate_Hourly_Temp2m.py file in the folder dataset_management in this repo
Hourly Surface Pressure	Jan 01 2017 - Dec 31 2020	hourly	35064	The hourly_surface_pressure file in https://code.earthengine.google.com/?accept_repo=users/cassandralem/tigertail-data and the Generate_Hourly_Surface_Pressure.py file in the folder dataset_management in this repo
Hourly Soil Temperature (Level 1)	Jan 01 2017 - Dec 31 2020	hourly	35064	The hourly_soil_temp_1 file in https://code.earthengine.google.com/?accept_repo=users/cassandralem/tigertail-data and the Generate_Hourly_Soil_Temp_1.py file in the folder dataset_management in this repo
Hourly Soil Temperature (Level 2)	Jan 01 2017 - Dec 31 2020	hourly	35064	The hourly_soil_temp_2 file in https://code.earthengine.google.com/?accept_repo=users/cassandralem/tigertail-data and the Generate_Hourly_Soil_Temp_2.py file in the folder dataset_management in this repo
Hourly Soil Temperature (Level 3)	Jan 01 2017 - Dec 31 2020	hourly	35064	The hourly_soil_temp_3 file in https://code.earthengine.google.com/?accept_repo=users/cassandralem/tigertail-data and the Generate_Hourly_Soil_Temp_3.py file in the folder dataset_management in this repo
Hourly Soil Temperature (Level 4)	Jan 01 2017 - Dec 31 2020	hourly	35064	The hourly_soil_temp_4 file in https://code.earthengine.google.com/?accept_repo=users/cassandralem/tigertail-data and the Generate_Hourly_Soil_Temp_4.py file in the folder dataset_management in this repo
Hourly Volumetric Soil Water (Layer 1)	Jan 01 2017 - Dec 31 2020	hourly	35064	The hourly_volu_soil_water_1 file in https://code.earthengine.google.com/?accept_repo=users/cassandralem/tigertail-data and the Generate_Hourly_Volu_Soil_Water_Layer1.py file in the folder dataset_management in this repo
Hourly Volumetric Soil Water (Layer 2)	Jan 01 2017 - Dec 31 2020	hourly	35064	The hourly_volu_soil_water_2 file in https://code.earthengine.google.com/?accept_repo=users/cassandralem/tigertail-data and the Generate_Hourly_Volu_Soil_Water_Layer2.py file in the folder dataset_management in this repo
Hourly Volumetric Soil Water (Layer 3)	Jan 01 2017 - Dec 31 2020	hourly	35064	The hourly_volu_soil_water_3 file in https://code.earthengine.google.com/?accept_repo=users/cassandralem/tigertail-data and the Generate_Hourly_Volu_Soil_Water_Layer3.py file in the folder dataset_management in this repo
Hourly Volumetric Soil Water (Layer 4)	Jan 01 2017 - Dec 31 2020	hourly	35064	The hourly_volu_soil_water_4 file in https://code.earthengine.google.com/?accept_repo=users/cassandralem/tigertail-data and the Generate_Hourly_Volu_Soil_Water_Layer4.py file in the folder dataset_management in this repo
BA1D Pressure and Temperature	1/18/19 11:15 - 2/29/20 17:21	every 15 minutes	39027	logged from BA1D directly
Oman tidal data set	2018-04-08 00:00:00 - 2020-09-04 00:45:00+00:00	every 15 minutes	844884	from sohn and matter 2023

Bubbles

This is the dataset from Aiken et al., 2022 for the bubbles detected in BA1B.

It was constructed by using a detected bubble as a template and then used a matched filter template matching algorithm to find the other bubbles.

It has the following columns:

1. time - a datetime column for the detection time of the bubble
2. similarity - the cross correlation similarity with the origin bubble event
3. template_id - the template ID for the detected events, there was only one bubble template so this column is always 0
4. ones - a column of 1s to make a cumulative count plot easy to make

Daily Precipitation

This data set was constructed from ERA5-land daily precipitation data and the hydrobasins catchement shape files.

It has the following columns:

1. datetime - date for precipitation
2. total_precipitation_sum - the total accumulated precipitation for the day

Geology

This data set is created from the AI framework paper, Aiken et al., In review.

For a full description please see the paper.

It has the following columns:

1. CORE - the core piece number
2. SECTION - the section of the core
3. Cell abundance (cells/g) - the amount of cells detected in the core
4. Mean dry electrical Resistivity (ohmm)
5. Bulk density (g/cm³) - this is a useful column for the amount of peridotite alteration present
6. AMS bulk susceptibility
7. LOI wt%
8. CO2 wt%
9. H20 wt%
10. CaCO3 calc
11. SECTION_UNIT
12. % of fractures - this is a calculated column based on the number of pixels labeled in the data
13. IMAGES
14. SEGMENTATION
15. TOP_DEPTH - the depth at the top of the core section when taken
16. ALTERATION
17. REMARKS1 REMARKS2 REMARKS4 REMARKS5 - the raw text remarks used for keyword generation
18. PnS2_sum PnL_sum PnP3V_sum PnP3H_sum PnP4_sum PnP6V_sum FnS2_sum FnL_sum FnP3V_sum FnP3H_sum FnP4_sum FnP6V_sum - the calculated connectivity statistics based on different polytopes
19. UNIT_TYPE_Dunite UNIT_TYPE_Fault rock UNIT_TYPE_Gabbro UNIT_TYPE_Harzburgite UNIT_TYPE_Metagabbro UNIT_TYPE_Other UNIT_CLASS_OPHIO UNIT_CLASS_UND
20. TEXTURES_Brecciated TEXTURES_Sheared
21. GRAINSIZE_Cryptocrystalline GRAINSIZE_Fine grained GRAINSIZE_Medium grained GRAINSIZE_Microcrystalline GRAINSIZE2_Coarse grained GRAINSIZE2_Cryptocrystalline GRAINSIZE2_Fine grained GRAINSIZE2_Medium grained GRAINSIZE2_Pegmatitic
22. Alteration_dummies_50%-90% Alteration_dummies_>90% - columns indicating the amount of alteration according to field geologists
23. Veins Serpentine vein Oxidation Carbonate veins Network Dyke Black serpentinization White veins Open cracks Dunite Gabbro Microgabbro Green veins Open crack Irregular Waxy green Alteration Subvertical Fine grained Subhorizontal Lineation Magnetite Thickness Harzburgite Altered gabbro Offset Altered Crack Pxenites Microbio sample Bulk serp Bulk Coalescence Waxy Wavy Slickensides Alteration halo Plagioclase Fracture Sheared Pyroxenite Striations Branching Blue patches Magmatic intrusions Hydrothermal Rodingite Magmatic veins Offsets Shearing Dark green Dunitic zone SiO2 TiO2 Al2O3 Fe2O3t MnO MgO CaO Na2O K2O P2O5 100*Fe(III)/FeT Vrecal Crrecal Co Nirecal Curecal Znrecal Srrecal - columns indicating this keyword was selected by chatgpt for this section
24. Redness Greenness Blueness Y (luminance) - columns related to the image itself

Hourly Precipitation

This data set was constructed from ERA5-land hourly precipitation data and the hydrobasins catchment shape files.

It has the following columns:

1. datetime - timestamp in the format YYYY-MM-DD HH:mm:ss.
2. total_precip - total precipitation amount during that hour, in meters (m)

Hourly Temperature 2m

This data set was constructed from ERA5-land hourly temperature data and the hydrobasins catchment shape files.

It has the following columns:

1. datetime - timestamp in the format YYYY-MM-DD HH:mm:ss.
2. temperature_2m - temperature 2m above the surface during that hour, in degrees Kelvin (K).

Hourly Surface Pressure

This data set was constructed from ERA5-land hourly surface pressure data and the hydrobasins catchment shape files.

It has the following columns:

1. datetime - timestamp in the format YYYY-MM-DD HH:mm:ss.
2. surface_pressure - pressure of the atmosphere on the surface of the land/sea/in-land water during that hour, in Pascals (Pa).

Hourly Soil Temperature (Levels 1, 2, 3, 4)

These data sets were constructed from ERA5-land hourly soil temperature data and the hydrobasins catchment shape files.

Each one has the following columns:

1. datetime - timestamp in the format YYYY-MM-DD HH:mm:ss.
2. soil_temperature_level_# - temperature of the soil in level # of the ECMWF Integrated Forecasting System during that hour, in degrees Kelvin (K)

Hourly Volumetric Soil Water (Layers 1, 2, 3, 4)

These data sets were constructed from ERA5-land hourly volumetric soil water data and the hydrobasins catchment shape files.

Each one has the following columns:

1. datetime - timestamp in the format YYYY-MM-DD HH:mm:ss.
2. volumetric_soil_water_layer_# - volume fraction of water in the soil in layer # of the ECMWF Integrated Forecasting System during that hour

BA1D Pressure and Temperature

This data set was logged from BA1D directly.

It has the following columns:

1. datetime - timestamp UTC
2. elapsed_sec - the time in elapsed seconds from the start time
3. pressure_bar - the pressure in bars in the borehole fluid pressure
4. temperature_c - the temperature in celsius in the borehole fluid

Oman Tides

This data was calculated from water level changes in BA1D and come from Sohn and Matter, 2023.

1. datetime - timestamp UTC
2. tide_nstr - areal strain (unitless)