Total Carbon Dioxide Emissions From All Sectors, Coal for United States (EMISSCO2TOTVTTCOUSA) | FRED | St. Louis Fed
- Rows
- 52
- Columns
- 2
- Data range
- 1970-01-01 to 2021-01-01
Graph and download economic data for Total Carbon Dioxide Emissions From All Sectors, Coal for United States (EMISSCO2TOTVTTCOUSA) from 1970 to 2021 about coal, carbon dioxide emissions, sector, and USA.
AI analysis
Data Quality & Structural Analysis: US Coal CO2 Emissions (EMISSCO2TOTVTTCOUSA)
1. Dataset Overview & Value
This dataset, sourced from FRED (Federal Reserve Bank of St. Louis) via their public API (fredgraph.csv?id=EMISSCO2TOTVTTCOUSA), tracks annual total carbon dioxide emissions from coal combustion across all sectors in the United States, spanning 1970–2021. As a FRED-hosted series, it benefits from institutional-grade curation and standardized formatting, making it highly reliable for merging with other macroeconomic or environmental time series. The annual granularity (52 observations, one per year) and long historical span make this a strong candidate for long-run trend analysis, energy transition studies, and macro-environmental correlation work — particularly examining how coal-based emissions relate to GDP growth, energy prices, or policy interventions (e.g., Clean Air Act amendments, Paris Agreement post-2015).
2. Data Quality Observations
Data quality is excellent on the surface metrics: zero null cells across both columns, and 52 distinct values in each column matching the 52 total rows — indicating no null-driven distortion and no repeated observation dates (a good sign for time-series integrity). Duplicate-row computation is flagged as pending ("Phase D — recomputed against Parquet"), so uniqueness should be re-verified once that check completes, but the distinct-count match (52 distinct dates = 52 rows) is a strong proxy indicator that duplicates are unlikely. One caveat: the "sample data" rows are not yet available, so I cannot visually confirm formatting consistency (e.g., date parsing, decimal precision) — this should be validated before production use.
3. Key Column Distributions
The core metric column, EMISSCO2TOTVTTCOUSA, ranges from 877.93 to 2,175.42 (units likely million metric tons CO2), with a mean of 1,656.36 and median of 1,709.62. The mean-median gap combined with a skewness of -0.263 suggests a mild left skew — consistent with a series that rose through the mid-20th century, peaked, and has been declining more recently (post-2005 coal decline in the US is well documented), pulling the tail toward lower values. The interquartile range (Q1=1,319.57, Q3=1,987.82) spans roughly 668 units, indicating substantial variance (σ=367.32) relative to the mean — this is a volatile, structurally-shifting series rather than a stable trend, which is important context for correlation modeling (consider first-differencing or log transforms rather than levels).
4. Recommended Join Keys
The clear join key is observationdate (Date type, 0 nulls, 52 distinct values covering 1970–2021 annually). This aligns naturally with other FRED annual series and most US-based macroeconomic/environmental datasets that use calendar-year timestamps. When joining with monthly or quarterly datasets, aggregate (mean/sum) to annual frequency first to preserve alignment.
5. Suggested Complementary Datasets
For meaningful correlation discovery, pair this with: (1) other FRED emissions series by sector/fuel (e.g., petroleum or natural gas CO2 emissions, EMISSCO2TOTVTTCOUSA siblings) to decompose total US emissions trends; (2) US GDP or industrial production indices to test emissions-intensity/decoupling hypotheses; (3) coal production and price data (e.g., EIA coal consumption series) to link upstream supply with downstream emissions; (4) energy policy/regulatory event datasets for structural break analysis; and (5) renewable energy capacity growth series to quantify substitution effects as coal's share declines.
Columns
- observation_date (date)
- EMISSCO2TOTVTTCOUSA (decimal)