Sea ice records over more than a century at an observatory facing the Okhotsk coast of Hokkaido, Japan

Time series of sea ice dates and amounts at Abashiri.

The dataset contains four directories:

raw_data, csv_data, codes, and figs.

raw_data directory contains manually created data in an ascii format:

Si_dates_1892-2025.txt: yearly first and last sea ice dates (FD_sight, FD_shore, LD_sight, and LD_shore) during 1892-2025.

Total_ice_1956-2021.txt: daily total ice amounts during 1956-2021.

Drift_ice_1930-2021.txt: daily drift ice amounts during 1930-2021.

In Total_ice_1956-2021.txt and Drift_ice_1930-2021.txt, six monthly lines with daily values from previous December to May are described for each year; for the daily data, 0-9 denote the area fraction in quantile, * denotes full ice cover (10), + and - denote 0+ and 10-, respectively.

csv_data directory contains data files in a csv format:

Si_dates_1892-2025.csv, Total_ice_1956-2021.csv, and Drift_ice_1930-2021.csv.

Dates are described in an ISO 8601 format ("YYYY-MM-DD"; "0000-00-00" denotes date not defined; "9999-99-99" denotes missing observation). For total/drift ice amounts, *, -, and + are converted to 10.0, 9.75, and 0.25, respectively.

codes directory contains shared codes:

txt2csv_dates.py converts Si_dates_1892-2025.txt to Si_dates_1982-2025.csv.

txt2csv_ice.py converts ice amount files (.txt to .csv).

plot_fig6a.py and plot_fig7a.py plot ice amount time series (Figs. 6a and 7a) from the csv files.

cumul_dev_si_dates.F90 (or cumul_dev_si_dates.py) calculates cumulative deviation from the mean (CDM) values, which are written as cumul_dev_si_dates.dat (cumul_dev_si_dates.csv).
cumul_dev_si_dates_plot.gs (plot_cumul_dev_si_dates.py) draws a plot for the CDM time series.

figs directory contains figures.