Our World in Data - Compare CO2 Emissions per Person for the US, China and India, with the Source Citation (public chart API, no key)

import csv, io, json, time
 
SLUG = "co-emissions-per-capita"  # old slug co2-emissions-per-capita answers 301 to this one; fetch() follows it
WANT = {"USA": "United States", "CHN": "China", "IND": "India"}  # ISO 3166-1 alpha-3 codes as OWID uses them
CONTEXT = ["OWID_WRL"]  # World, for scale; drop it if you only want the three countries
 
SITE = "https://ourworldindata.org"
PAUSE = 1.0  # no published rate limit; keep calls serial and spaced
 
def on_owid_tab():
    """fetch() runs same-origin from a logged-out ourworldindata.org tab (/robots.txt, plain text). Reuse it."""
    try:
        if js("location.origin") == SITE:
            return
    except Exception:
        pass
    new_tab(SITE + "/robots.txt")
    wait_for_load()
    body = js("document.body ? document.body.innerText : ''") or ""
    if js("location.origin") != SITE or "Sitemap:" not in body:
        raise RuntimeError(f"ourworldindata.org did not serve /robots.txt (at {js('location.href')}): stop")
 
# js() gives up after a few seconds, so start the fetch, then poll for it.
START_JS = """(() => { window.__ow = window.__ow || {}; const id = String(Math.random()).slice(2); window.__ow[id] = null;
  fetch(%s, {credentials: 'omit', redirect: 'follow'})
    .then(async r => { window.__ow[id] = {status: r.status, url: r.url, redirected: r.redirected,
                                          type: r.headers.get('content-type'), text: await r.text()}; })
    .catch(e => { window.__ow[id] = {status: 0, url: '', redirected: false, type: '', text: String(e)}; });
  return id; })()"""
POLL_JS = """(() => { const v = window.__ow[%s]; if (v) delete window.__ow[%s]; return v; })()"""
 
def get(path, kind, timeout=60):
    """GET SITE+path (or an absolute https URL) from inside the tab, following 301s (slugs get renamed). kind is 'csv' or 'json'.
    Unknown slug: HTTP 404 {"status":404,"error":"Not found"}. Waits 30 s once on 429/5xx, raises on anything else."""
    for attempt in range(2):
        rid = json.dumps(js(START_JS % json.dumps(path if path.startswith("https://") else SITE + path)))
        t0 = time.time()
        while (r := js(POLL_JS % (rid, rid))) is None:
            if time.time() - t0 > timeout:
                raise RuntimeError(f"{path}: no answer after {timeout}s")
            time.sleep(0.3)
        time.sleep(PAUSE)
        if r["status"] in (0, 429, 500, 502, 503, 504) and attempt == 0:
            time.sleep(30)
            continue
        if r["status"] != 200 or kind not in (r["type"] or ""):
            raise RuntimeError(f"{path}: HTTP {r['status']} {r['type']} {' '.join(r['text'].split())[:160]!r}; stopping")
        return r
 
on_owid_tab()
 
# 1. The latest year for the three countries (plus World). csvType=filtered is what makes country= and time= apply.
codes = list(WANT) + CONTEXT
q = f"v=1&csvType=filtered&useColumnShortNames=true&country={'~'.join(codes)}&time=latest"
r = get(f"/grapher/{SLUG}.csv?{q}", "csv")
final_slug = r["url"].split("/grapher/")[1].split(".csv")[0]
if r["redirected"]:
    print(f"slug {SLUG!r} redirected to {final_slug!r}: use the new one")
rows = list(csv.DictReader(io.StringIO(r["text"])))
head = list(rows[0].keys()) if rows else []
if head[:3] != ["entity", "code", "year"] or len(head) != 4:
    raise RuntimeError(f"unexpected CSV header {head}: re-check the chart (it may have several indicators)")
col = head[3]
got = {x["code"]: x for x in rows}
missing = [c for c in codes if c not in got]
if missing:  # unknown codes and entities without data in the latest year are dropped silently, with HTTP 200
    raise RuntimeError(f"no row for {missing}: check the code, or ask for time=earliest..latest and take each country's last year")
years = {got[c]["year"] for c in WANT}
print(f"{len(rows)} rows, column {col!r}, year(s) {sorted(years)}")
 
# 2. Title, unit and citation from the same chart's metadata.
m = json.loads(get(f"/grapher/{final_slug}.metadata.json?v=1&useColumnShortNames=true", "json")["text"])  # same key as the CSV column
c = m["columns"].get(col)
if not c or not c.get("citationShort"):
    raise RuntimeError(f"metadata has no citation for column {col!r} (columns: {list(m['columns'])}): stop")
print(f"chart: {m['chart']['title']!r}, unit {c['unit']!r}, timespan {c['timespan']}, "
      f"lastUpdated {c['lastUpdated']}, nextUpdate {c['nextUpdate']}")
print(f"what it counts: {c['descriptionShort']}")
 
# 3. The comparison.
print()
for code in sorted(codes, key=lambda k: -float(got[k][col])):
    x = got[code]
    print(f"  {x['entity']:<14} {x['code']:<8} {x['year']}  {float(x[col]):6.2f} {c['shortUnit']}")
val = {k: float(got[k][col]) for k in WANT}
ks = sorted(WANT, key=lambda k: -val[k])  # works for any WANT, not just these three
print("\n  " + ", ".join(f"{WANT[a]}/{WANT[b]} {val[a] / val[b]:.2f}x" for i, a in enumerate(ks) for b in ks[i + 1:]))
 
# 4. Who made the underlying data, and under which licence (the FAQ: respect the providers' licences and cite them).
full = json.loads(get(c["fullMetadata"], "json")["text"])  # api.ourworldindata.org, CORS open
if full.get("nonRedistributable"):
    raise RuntimeError("this indicator is marked nonRedistributable: do not republish the numbers")
print("\nproviders:")
for o in full.get("origins", []):
    lic = o.get("license") or {}
    print(f"  {o['producer']} - {o['title']} (published {o.get('datePublished')}): {lic.get('name')} {lic.get('url')}")
 
year = years.pop() if len(years) == 1 else "/".join(sorted(years))
print(f"\nanswer: in {year}, CO2 emissions per person were " + ", ".join(f"{WANT[k]} {val[k]:.1f} t" for k in ks)
      + f" ({c['unit']}, fossil fuels and industry, excluding land-use change).")
print(f"  Source: {c['citationShort']}. Our World in Data, \"{m['chart']['title']}\", "
      f"https://ourworldindata.org/grapher/{final_slug} (CC BY 4.0), retrieved {m['dateDownloaded']}.")

4 rows, column 'emissions_total_per_capita', year(s) ['2025']
chart: 'CO₂ emissions per capita', unit 'tonnes per person', timespan 1750-2025, lastUpdated 2026-11-10, nextUpdate 2027-11-10
what it counts: Carbon dioxide (CO₂) emissions from burning fossil fuels and industrial processes. This includes emissions from transport, electricity generation, and heating, but not land-use change.
 
  United States  USA      2025   14.50 t/person
  China          CHN      2025    8.77 t/person
  World          OWID_WRL 2025    4.74 t/person
  India          IND      2025    2.19 t/person
 
  United States/China 1.65x, United States/India 6.63x, China/India 4.01x
 
providers:
  Global Carbon Project - Global Carbon Budget (published 2026-11-10): CC BY 4.0 https://doi.org/10.5281/zenodo.5569234
  Global Carbon Project - Global Carbon Budget (published 2026-11-10): CC BY 4.0 https://www.icos-cp.eu/data-services/about-data-portal/data-license
  Various sources - Population (published 2024-07-15): CC BY 4.0 https://creativecommons.org/licenses/by/4.0/
 
answer: in 2025, CO2 emissions per person were United States 14.5 t, China 8.8 t, India 2.2 t (tonnes per person, fossil fuels and industry, excluding land-use change).
  Source: Global Carbon Budget (2026); Population based on various sources (2024) – with major processing by Our World in Data. Our World in Data, "CO₂ emissions per capita", https://ourworldindata.org/grapher/co-emissions-per-capita (CC BY 4.0), retrieved 2026-11-13.

import json, time
 
SITE = "https://ourworldindata.org"
 
# robots.txt and the FAQ answers on reuse, as the pages show them.
new_tab(SITE + "/robots.txt")
wait_for_load()
if js("location.origin") != SITE:
    raise RuntimeError(f"not on {SITE} (at {js('location.href')}): stop")
print("robots.txt:", repr(js("document.body.innerText").strip()))
 
goto_url(SITE + "/faqs")
wait_for_load()
text = " ".join((js("document.body.innerText") or "").split())
if "Do you have an API?" not in text:
    raise RuntimeError(f"the FAQ page did not load (at {js('location.href')}): stop")
for q in ("For users who work with automated workflows",
          "Most of the data on Our World in Data comes from third-party providers",
          "You are only permitted to use the Our World in Data name or logo",
          "However, the Grapher is not freely licensed"):
    i = text.find(q)
    if i < 0:
        raise RuntimeError(f"FAQ changed, {q!r} not found: re-read {SITE}/faqs before relying on this skill")
    print(f"\nfaq: {text[i:i + 260].rstrip()}")
 
# Same-origin probes, same tab. No retries: this block only inspects answers.
goto_url(SITE + "/robots.txt")
wait_for_load()
START_JS = """(() => { window.__ow = window.__ow || {}; const id = String(Math.random()).slice(2); window.__ow[id] = null;
  fetch(%s, {credentials: 'omit'})
    .then(async r => { window.__ow[id] = {status: r.status, url: r.url, redirected: r.redirected,
                                          type: r.headers.get('content-type'), text: await r.text()}; })
    .catch(e => { window.__ow[id] = {status: 0, url: '', redirected: false, type: '', text: String(e)}; });
  return id; })()"""
POLL_JS = """(() => { const v = window.__ow[%s]; if (v) delete window.__ow[%s]; return v; })()"""
 
def raw(path, timeout=60):
    rid = json.dumps(js(START_JS % json.dumps(SITE + path)))
    t0 = time.time()
    while (r := js(POLL_JS % (rid, rid))) is None:
        if time.time() - t0 > timeout:
            raise RuntimeError(f"{path}: no answer after {timeout}s")
        time.sleep(0.3)
    time.sleep(1.0)
    return r
 
F = "v=1&csvType=filtered&useColumnShortNames=true"
print()
for path in (f"/grapher/co2-emissions-per-capita.csv?{F}&country=USA~CHN~IND&time=latest",   # renamed slug: 301
             f"/grapher/no-such-chart-xyz.csv?{F}&country=USA&time=latest",                  # unknown slug
             f"/grapher/co-emissions-per-capita.csv?{F}&country=USA~XYZ&time=latest",         # unknown code: dropped
             f"/grapher/co-emissions-per-capita.csv?{F}&country=United%20States~China&time=latest",  # names work too
             f"/grapher/co-emissions-per-capita.csv?{F}&country=USA&time=2099",               # future year: clamped
             f"/grapher/co-emissions-per-capita.csv?{F}&country=USA&time=2023..latest",
             f"/grapher/life-expectancy.csv?{F}&country=JPN~OWID_USS&time=latest",            # USSR ends 1913: dropped
             "/grapher/co-emissions-per-capita.csv?v=1&country=USA&time=latest"):            # no csvType: whole table
    r = raw(path)
    lines = r["text"].strip().split("\n")
    where = f" -> {r['url'].split('.org')[1].split('?')[0]}" if r["redirected"] else ""
    print(f"{path}\n  HTTP {r['status']} {r['type']}{where}, {len(lines)} lines: {' | '.join(lines[:3])[:150]}")
 
r = raw("/grapher/co-emissions-per-capita.metadata.json?v=1")
print(f"\nmetadata without useColumnShortNames: columns {list(json.loads(r['text'])['columns'])}")

robots.txt: 'Sitemap: https://ourworldindata.org/sitemap.xml'
 
faq: For users who work with automated workflows, computational notebooks, or custom applications, we offer direct URLs to access data in CSV format and comprehensive metadata in JSON format. You can fetch the complete data or only the subset of the data currently
 
faq: Most of the data on Our World in Data comes from third-party providers (such as the WHO, UN, and World Bank) and is subject to the license terms of those providers. You should always check their license before reusing or republishing the data. Our work would n
 
faq: You are only permitted to use the Our World in Data name or logo as it appears on our charts, articles, and other pages. You have this permission only as long as your use of the material we make available is in accordance with the license terms and follows the
 
faq: However, the Grapher is not freely licensed for reuse without permission. If you'd like to copy, modify, distribute, or use any part of the Grapher codebase in your own project, product, or service, you'll need to first request written permission from Global C
 
/grapher/co2-emissions-per-capita.csv?v=1&csvType=filtered&useColumnShortNames=true&country=USA~CHN~IND&time=latest
  HTTP 200 text/csv -> /grapher/co-emissions-per-capita.csv, 4 lines: entity,code,year,emissions_total_per_capita | China,CHN,2025,8.768821 | India,IND,2025,2.1883202
/grapher/no-such-chart-xyz.csv?v=1&csvType=filtered&useColumnShortNames=true&country=USA&time=latest
  HTTP 404 application/json; charset=utf-8, 1 lines: {"status":404,"error":"Not found"}
/grapher/co-emissions-per-capita.csv?v=1&csvType=filtered&useColumnShortNames=true&country=USA~XYZ&time=latest
  HTTP 200 text/csv, 2 lines: entity,code,year,emissions_total_per_capita | United States,USA,2025,14.500179
/grapher/co-emissions-per-capita.csv?v=1&csvType=filtered&useColumnShortNames=true&country=United%20States~China&time=latest
  HTTP 200 text/csv, 3 lines: entity,code,year,emissions_total_per_capita | China,CHN,2025,8.768821 | United States,USA,2025,14.500179
/grapher/co-emissions-per-capita.csv?v=1&csvType=filtered&useColumnShortNames=true&country=USA&time=2099
  HTTP 200 text/csv, 2 lines: entity,code,year,emissions_total_per_capita | United States,USA,2025,14.500179
/grapher/co-emissions-per-capita.csv?v=1&csvType=filtered&useColumnShortNames=true&country=USA&time=2023..latest
  HTTP 200 text/csv, 4 lines: entity,code,year,emissions_total_per_capita | United States,USA,2023,14.335961 | United States,USA,2024,14.21929
/grapher/life-expectancy.csv?v=1&csvType=filtered&useColumnShortNames=true&country=JPN~OWID_USS&time=latest
  HTTP 200 text/csv, 2 lines: entity,code,year,life_expectancy_0 | Japan,JPN,2023,84.7123
/grapher/co-emissions-per-capita.csv?v=1&country=USA&time=latest
  HTTP 200 text/csv, 27239 lines: Entity,Code,Year,CO₂ emissions per capita | Afghanistan,AFG,1949,0.0019921463 | Afghanistan,AFG,1950,0.010366015
 
metadata without useColumnShortNames: columns ['Annual CO₂ emissions (per capita)']

import csv, io, json, sys, time, urllib.error, urllib.parse, urllib.request
 
SLUG, CODES = "co-emissions-per-capita", ["USA", "CHN", "IND"]
BASE = "https://ourworldindata.org/grapher/"
# Cloudflare answers HTTP 403 "error code: 1010" to Python's default User-Agent (Python-urllib/3.x) unless the
# response is already cached. Name your client honestly; do not pretend to be a browser.
UA = {"User-Agent": "co2-compare-script/1.0 (+https://example.org/contact)"}
 
def get(url):
    """GET -> (final URL, body text). urllib follows the 301 of a renamed slug. Waits 30 s once on 429/5xx."""
    for attempt in range(2):
        try:
            with urllib.request.urlopen(urllib.request.Request(url, headers=UA), timeout=60) as r:
                out = r.geturl(), r.read().decode("utf-8")
            time.sleep(1)  # no published limit; stay serial
            return out
        except urllib.error.HTTPError as e:
            if e.code in (429, 500, 502, 503, 504) and attempt == 0:
                time.sleep(30)
                continue
            sys.exit(f"{url}: HTTP {e.code} {e.read().decode()[:160]}")  # unknown slug: 404; blocked UA: 403 "error code: 1010"
 
q = urllib.parse.urlencode({"v": 1, "csvType": "filtered", "useColumnShortNames": "true",
                            "country": "~".join(CODES), "time": "latest"}, safe="~")
final, text = get(f"{BASE}{SLUG}.csv?{q}")
slug = final.split("/grapher/")[1].split(".csv")[0]
rows = {r["code"]: r for r in csv.DictReader(io.StringIO(text))}
if sorted(rows) != sorted(CODES):  # unknown codes and countries without the latest year vanish with HTTP 200
    sys.exit(f"got {sorted(rows)}, wanted {sorted(CODES)}: stop")
col = [k for k in next(iter(rows.values())) if k not in ("entity", "code", "year")]
if len(col) != 1:
    sys.exit(f"expected one value column, got {col}")
col = col[0]
meta = json.loads(get(f"{BASE}{slug}.metadata.json?v=1&useColumnShortNames=true")[1])
c = meta["columns"][col]
for code in CODES:
    r = rows[code]
    print(f"{r['entity']:<14} {r['year']}  {float(r[col]):5.2f} {c['shortUnit']}")
print(f"Source: {c['citationShort']}. Our World in Data, https://ourworldindata.org/grapher/{slug} (CC BY 4.0), "
      f"data updated {c['lastUpdated']}.")

United States  2025  14.50 t/person
China          2025   8.77 t/person
India          2025   2.19 t/person
Source: Global Carbon Budget (2026); Population based on various sources (2024) – with major processing by Our World in Data. Our World in Data, https://ourworldindata.org/grapher/co-emissions-per-capita (CC BY 4.0), data updated 2026-11-10.

B=https://ourworldindata.org/grapher
Q='v=1&csvType=filtered&useColumnShortNames=true&country=USA~CHN~IND&time=latest'
R='--max-time 60 --retry 1 --retry-delay 30'  # like blocks 1 and 3: one retry after 30 s on 429/5xx (curl also retries a timeout)
curl -sS --fail -L $R "$B/co-emissions-per-capita.csv?$Q"; echo   # the CSV has no trailing newline
sleep 1
curl -sS --fail -L $R "$B/co-emissions-per-capita.metadata.json?v=1&useColumnShortNames=true" | python3 -c '
import json, sys
c = json.load(sys.stdin)["columns"]["emissions_total_per_capita"]
print(c["unit"], "| updated", c["lastUpdated"], "| Source:", c["citationShort"])
'
sleep 1
curl -sS --max-time 60 -o /dev/null -w 'old slug: HTTP %{http_code} -> %{redirect_url}\n' "$B/co2-emissions-per-capita.csv?$Q"

entity,code,year,emissions_total_per_capita
China,CHN,2025,8.768821
India,IND,2025,2.1883202
United States,USA,2025,14.500179
tonnes per person | updated 2026-11-10 | Source: Global Carbon Budget (2026); Population based on various sources (2024) – with major processing by Our World in Data
old slug: HTTP 301 -> https://ourworldindata.org/grapher/co-emissions-per-capita.csv?v=1&csvType=filtered&useColumnShortNames=true&country=USA%7ECHN%7EIND&time=latest