Atmospheric CO2 concentrations (Mauna Loa)
- Rows
- 823
- Columns
- 6
- Data range
- 1958-03-01 to 2026-09-01
- Auto-refresh
- Re-checks its source every 14 days
- Last refreshed with new data
- 2026-11-05
Monthly mean carbon dioxide measured at Mauna Loa Observatory, Hawaii, the longest continuous record of atmospheric CO2.
AI analysis
Atmospheric CO2 Concentrations (Mauna Loa) — Dataset Analysis
1. Dataset Overview & Value This dataset captures the Keeling Curve, the longest continuous instrumental record of atmospheric CO2, sourced from [datahub.io/core/co2-ppm](https://datahub.io/core/co2-ppm/r/co2-mm-mlo.csv), a curated open data publisher that maintains regularly refreshed core datasets. Spanning from 1958 through early 2026 (816 monthly observations), it represents one of the most scientifically significant time series in climate science. Its temporal depth and regularity make it exceptionally well-suited for long-run trend analysis, seasonal decomposition, and cross-domain correlation studies linking CO2 growth to industrial activity, policy events, and ecological outcomes.
2. Data Quality Observations Data quality here is notably strong. The total null cell count is zero across all 816 rows and 6 columns, which is rare for a dataset spanning nearly seven decades and suggests rigorous curation by the Mauna Loa Observatory and datahub.io. No type mismatches are flagged. The one meaningful quality concern lies in the "Number of Days" column, which carries 204 outlier values — roughly 25% of all rows — and a suspicious minimum of -9.99 alongside a negative mean of -2.01. This almost certainly reflects a sentinel/fill value encoding (e.g., -9.99 signaling missing or estimated day-count data) rather than a true measurement, and the column should be treated with caution or excluded from numeric analysis. Duplicate row status is pending Parquet reconciliation but is not expected to be an issue given the distinct=816 count on the Date column.
3. Key Column Distributions The Average CO2 column tells the core story: a mean of 360.85 ppm, median of 356.08 ppm, and a range of 312.42 to 430.51 ppm — a ~118 ppm rise over 68 years. The positive skew of 0.377 reflects accelerating growth in recent decades pulling the upper tail. The Interpolated column is nearly identical (mean 360.85, skew 0.379), confirming it serves as a gap-filled twin to Average, with slightly fewer distinct values (784 vs. 793), indicating smoothing. The Trend column is the most anomalous structurally: its range of -1 to 31 with only 25 distinct integer values and a strong negative skew of -0.958 suggests it is not a raw CO2 measurement but likely a derived seasonal adjustment index or anomaly score, clustered toward higher values. The Decimal Date column is perfectly symmetric (skew ≈ 0), confirming uniform monthly sampling with no temporal gaps.
4. Recommended Join Key Columns The Date column (type: Date, 816 distinct values, zero nulls) is the unambiguous primary join key for any time-series merge — it provides clean monthly granularity from 1958–2026. Decimal Date offers a useful secondary numeric key for regression frameworks or datasets that encode time as a fractional year. For aggregated joins (e.g., annual or decadal studies), a derived Year field extracted from Date would broaden compatibility with datasets that lack monthly resolution.
5. Recommended Pairing Datasets for Correlation Discovery Several dataset categories would pair powerfully with this record: - Global Surface Temperature Anomalies (NASA GISS or NOAA) — the most natural pairing to quantify the CO2-temperature relationship over the same monthly timeline - Global Fossil Fuel CO2 Emissions (e.g., Global Carbon Project) — to correlate emission events, economic recessions (2008, 2020), or policy inflection points against concentration changes - Arctic/Antarctic Sea Ice Extent (NSIDC) — to explore lagged correlations between CO2 accumulation and cryosphere decline - Global Energy Consumption by Source (IEA or BP Statistical Review) — to link the transition toward renewables against the rate of CO2 increase - Oceanic pH / Ocean Acidification records — CO2 absorption by oceans is a direct chemical consequence, making this a near-deterministic correlation worth modeling
Columns
- Date (date)
- Decimal Date (decimal)
- Average (decimal)
- Interpolated (decimal)
- Trend (integer)
- Number of Days (decimal)