Global Extreme Weather and Climate Change Dashboard
Detection & Attribution: IPCC AR6 vs. This Site
IPCC AR6 (2021) stands as the most current, most rigorously reviewed formal climate assessment with a dedicated extreme-events chapter (WG1 Chapter 11) -- see the Methodology page for the underlying IPCC detection/attribution definitions this whole site applies. Unlike the sibling US dashboard's version of this page (which compares against AR6's US-regional Chapter 12 findings), this table compares against AR6's global-scope SPM and Chapter 11 findings, matching the dashboard's global scope. Each row's "GWX finding" column computes live from the exact same data as that hazard's phenomena page, so this page can never silently drift out of sync with what those pages show.
Detection only, not attribution. This page (and every page on this site) assesses whether a change has occurred, not why -- attribution (evaluating specific causal factors, e.g. how much of a change traces to human-caused warming) requires model-based analysis this project hasn't undertaken. See the Methodology page for the IPCC's definitions of the two terms.
What's New Since AR6
AR6 (2021) stands as the most current formal assessment, but the record and literature have moved since its cutoff. The list below presents evidence -- published post-2021, or computed directly on this site -- that complicates or narrows specific AR6 detection findings, not a comprehensive new assessment:
- Tropical cyclones, proportion of Category 3-5 storms (AR6 SPM A.3.4). Kossin et al.'s late-2020 correction -- published before AR6's citation deadline -- narrowed the statistically significant result from global to just 2 ocean basins (under 20% of global TC activity), a revision AR6's "likely... increased" language does not appear to carry forward. Independently, the NATL+WPAC robustness check below finds the %-major-hurricanes finding holds over 1980-2025 but does not survive extension to the full 1945-2025 record for those two basins -- see the Tropical Cyclones page and the row below.
- Wildfire, "fire weather" vs. realized fire. Jones et al. 2022 (a 500+-study post-AR6 synthesis) and the GWIS burned-area series tracked here both confirm global burned area has been declining even as fire weather conditions become more favorable in some regions -- AR6's fire-weather finding and the burned-area decline don't contradict each other, but public discussion often conflates the two. See the Wildfire page.
- Heat extremes, pace of the hottest days vs. the seasonal average. McKinnon, Simpson & Williams (2024, PNAS), using 90 years of station data, found the hottest summer days have warmed at essentially the same rate as the seasonal median -- the widening of the warm tail traces to the cold tail warming more slowly, not the hot tail accelerating. This nuances, rather than contradicts, AR6's frequency/intensity finding for heat extremes.
Read the "Consistency" column as a qualitative judgment call, not a formal statistical test: AR6's findings synthesize many studies and multiple lines of evidence globally, while the GWX findings in the table below come from single-site, live-computed metrics built on specific data sources. Where this site tracks several alternative metrics for one hazard, the table below shows every metric's verdict, not just whichever one happens to agree with AR6.
| Hazard | IPCC AR6 finding (global) | GWX finding | Consistency |
|---|---|---|---|
| Heat Waves | "It is virtually certain that hot extremes (including heatwaves) have become more frequent and more intense across most land regions since the 1950s." |
| Consistent The bespoke heat wave index and all three standard ETCCDI/CLIMDEX metrics (TXx, WSDI, TN90p) -- built from an entirely independent station network (GHCN-Daily's global GSN, not any US-specific dataset) -- all show a detected increase over their reliable windows (1949-present). A clean, direct match across four independently-constructed metrics on an independent global station network, not just one headline number. |
| Cold Extremes | "...cold extremes (including cold waves) have become less frequent and less severe [across most land regions since the 1950s], with high confidence that human-induced climate change is the main driver of these changes." |
| Consistent 3 of 4 cold-side ETCCDI/CLIMDEX metrics (CSDI, TN10p, Frost Days) show a detected decrease over their reliable window (1949-present) -- directly matching AR6's "less frequent and less severe" finding. The 4th (TNn, the annual coldest night) shows a sub-threshold trend (Mann-Kendall p=0.103, just above the <10% detection bar) -- a magnitude issue, matching AR6's direction. |
| Tropical Cyclones | "It is likely that the global proportion of Category 3-5 tropical cyclone instances has increased over the past four decades." |
| Partially consistent This finding traces to Kossin et al. (2020). Roger Pielke Jr.'s documented critique of how it was carried into AR6 raises three objections. First, the citation chain drifts semantically: operational storm "fixes" (individual satellite observations) become Kossin's "exceedances," then Ch.11's "instances," then the SPM's "occurrences" -- obscuring that the study measured observation density, not distinct-storm frequency. Second, Kossin et al.'s late-2020 correction, filed before AR6's citation deadline, narrowed the significant result from global to 2 ocean basins (under 20% of global TC activity), a narrowing AR6's SPM text doesn't carry through. Third, the paper itself states its results "do not constitute a traditional formal detection" and cannot precisely quantify anthropogenic contribution -- caveats the SPM's "likely... increased" framing drops. An independent check here lands in the same place: % of hurricanes that are major shows a detected change over 1980-2025 globally (p=0.03) and barely replicates in the NATL+WPAC-only subset over that window (p=0.053), but fails to survive the full 1945-2025 NATL+WPAC record (p=0.66); the Category-4-specific increase detected globally never replicates in that 2-basin subset at any window length. The signal holds only in the specific recent decades the literature examined, not the robust, basin-general, long-record result "likely increased" implies -- the same fragility this critique describes. Full robustness check: Tropical Cyclones. |
| Drought | "Human-induced climate change has contributed to increases in agricultural and ecological droughts in some regions due to increased land evapotranspiration (medium confidence)." |
| Consistent The SPEI-style series (precipitation minus atmospheric evaporative demand, AED) shows a detected increase in severe-or-worse drought area at every severity level (D0-D4); AED supplies the mechanism AR6's quote names ("increased land evapotranspiration"). The precipitation-only SPI-style series shows no detected change at any level, consistent with precipitation alone not driving the trend AR6 identifies. This divergence between the two series forms the Drought page's central finding, matching AR6's stated mechanism directly, not only the direction of change. |
| Wildfire | "Fire weather conditions (compound hot, dry and windy events) have become more probable in some regions (medium confidence)." |
| Not directly comparable This page tracks two quantities distinct from AR6's subject here, and under the 30-year minimum record-length standard, neither currently renders a detection verdict: GFED5's dry-matter-combusted record starts 2001 (its pre-MODIS 1997-2000 years are excluded as lower-quality, see the Wildfire page), only 23-24 years so far; GWIS's burned-area record starts 2002, only 22-23 years. Both clear the 30-year bar in the early 2030s. Earlier, shorter-window checks found GFED5's total showing no detected change while GWIS's global burned area showed a detected decline (matching Andela et al. 2017's documented decline in global burned area) -- valid results at the time, retired here under a stricter standard rather than left displayed on a record too short to trust. Neither quantity matches "fire weather conditions" (AR6's subject here, an atmospheric-favorability measure): both measure realized combustion and burned area, which also depend on fuel availability, land management, and suppression, beyond whether conditions favored fire. |
| Extratropical Cyclones / Winter Storms | "There is low confidence in past changes of maximum wind speeds and other measures of dynamical intensity of extratropical cyclones." Separately: "Mid-latitude storm tracks have likely shifted poleward in both hemispheres since the 1980s, with marked seasonality in trends (medium confidence)." |
| Consistent The annual most-intense-storm minimum pressure shows no detected change (p=0.49) -- matching AR6's low-confidence characterization of intensity trends specifically. This page doesn't track storm-track position, so it can't test the poleward-shift finding either way. |
| Flooding | "Significant trends in peak streamflow have been observed in some regions over the past decades (high confidence)... Regional changes in river floods are more uncertain than changes in pluvial floods because complex hydrological processes and forcings, including land cover change and human water management, are involved." |
| Consistent AR6 declines to state a clean global river-flood direction, citing regional inconsistency -- the GloFAS-based high-flow-cell index shows the same pattern: no detected change globally (the raw Mann-Kendall p=0.070 clears IPCC's flat 10% example threshold, but not the multiple-comparisons-adjusted bar for testing 57 variables at once -- see the Methodology page), which itself masks regional divergence underneath -- Africa and South America both show a detected decrease, while Asia, Europe, North America, and Oceania show no detected change at all. AR6's hedged wording anticipates this patchy, regionally inconsistent picture rather than a uniform trend. |
| Severe Convective Storms | "There is low confidence in past trends in characteristics of severe convective storms, such as hail and severe winds, beyond an increase in precipitation rates." |
| Not directly comparable US tornado counts (the one direct severe-convective-storm dataset available -- there is no unified global tornado database) split by severity: the all-tornado total's reliable window (1997-present) remains under the 30-year minimum and renders no detection verdict until 2027. Significant tornadoes (F/EF2+), far less exposed to the NEXRAD-era weak-tornado detection artifact that sets the total's window, run a longer, separately reliable window back to 1980 (added 2026-08-19) and show no detected change over that 46-year record -- an actual finding, not an insufficient-data non-result. Global CAPE, this page's proxy for atmospheric instability favorable to severe convection, is a longer, 1979-based record and DOES clear detection on all three of its series -- but not uniformly in one direction: mean CAPE shows a detected DECREASE, while high-CAPE-day counts show a detected INCREASE and widespread-instability-day counts a detected decrease. CAPE measures instability potential, not realized storm occurrence, and AR6's statement is specifically about storm characteristics (hail, wind), not the broader thermodynamic environment -- a mixed-direction, detected change in the ingredient doesn't mechanically imply any particular change in the outcome, so this isn't read as contradicting or confirming AR6's low-confidence framing, just measuring something related but distinct. |
IPCC AR6 Chapter 12 Regional Consistency — Africa
AR6's regional detail reports an observed increase in agricultural and ecological drought across most African sub-regions (West, Central, and Southern Africa) — matching the detected drought increase computed here for the continent. Flooding runs more mixed: several sub-regions report an observed increase in river or pluvial flooding, while the GloFAS-based index here shows a detected decrease in high-flow conditions for Africa overall — a divergence, not smoothed over. AR6's tropical-cyclone and fire-weather findings for Africa remain projections only, with no observed baseline to check the no-detected-change results here against.
Source: AR6 WGI Regional Fact Sheet — Africa — this site detects observed historical trends only, never projections, so this comparison draws only on AR6's observed findings, excluding its projected ones.
| Variable | GWX finding | AR6 Ch.12 consistency |
|---|---|---|
| Africa: % of land in D2+ (severe or worse) drought | computing… | Consistent AR6 reports observed increases in agricultural and ecological drought across West Southern, East Southern, Central, and West Africa (medium-high confidence), matching the detected drought increase here for the continent. |
| Africa: % of pour-point cells in high flow | computing… | Partially consistent AR6 reports observed increases in heavy precipitation and pluvial or river flooding in West Southern, East Southern, and West Africa, while the GloFAS-based index here shows a detected decrease in high-flow conditions for Africa overall -- a divergence, not smoothed over. |
| Africa: dry matter combusted | computing… | Not directly comparable AR6's Africa fact sheet has no observed (only projected) fire-weather finding to compare against, and the wildfire records here are currently under the 30-year minimum record length in any case. |
| Africa: burned area | computing… | Not directly comparable AR6's Africa fact sheet has no observed (only projected) fire-weather finding to compare against, and the wildfire records here are currently under the 30-year minimum record length in any case. |
| Africa: ACE (SIO) | computing… | Not directly comparable AR6's tropical-cyclone findings for Africa (East Southern Africa, Madagascar) are projections only, with no observed baseline to check the no-detected-change result here against. |
IPCC AR6 Chapter 12 Regional Consistency — Asia
AR6 reports observed drought increases in parts of East and Southwest Asia, and an observed increase in the intensity of strong tropical cyclones affecting East and Southeast Asia. The continent-wide indices tracked here show a sub-threshold drought trend and no detected change in tropical cyclone energy — plausibly diluted by aggregating across a large, climatically diverse continent, rather than contradicting the more localized AR6 findings. Fire weather (AR6: lengthening in North Asia) and realized burned area or combustion (tracked here: a detected decrease) measure different quantities, the same distinction drawn on the Wildfire page.
Source: AR6 WGI Regional Fact Sheet — Asia — this site detects observed historical trends only, never projections, so this comparison draws only on AR6's observed findings, excluding its projected ones.
| Variable | GWX finding | AR6 Ch.12 consistency |
|---|---|---|
| Asia: % of land in D2+ (severe or worse) drought | computing… | Partially consistent AR6 reports observed drought increases in much of continental East Asia and, since the 1980s, Southwest Asia, but the continent-wide index here shows only a sub-threshold trend (19% of variability) -- plausibly diluted by averaging across a very large, climatically diverse continent. |
| Asia: % of pour-point cells in high flow | computing… | Partially consistent AR6 reports an observed increase in flood-producing heavy precipitation in North Asia since the mid-1970s, while the continent-wide GloFAS index here shows no detected change -- plausibly diluted the same way as drought. |
| Asia: dry matter combusted | computing… | Not directly comparable AR6's Asia fact sheet frames fire-season lengthening as a projection, not an observed finding, and the wildfire records here are currently under the 30-year minimum record length regardless. |
| Asia: burned area | computing… | Not directly comparable AR6's Asia fact sheet frames fire-season lengthening as a projection, not an observed finding, and the wildfire records here are currently under the 30-year minimum record length regardless. |
| Asia: most extreme storm's minimum pressure (annual) | computing… | Not directly comparable AR6's Asia fact sheet has no observed extratropical-cyclone or storm-track finding to compare against. The metric tracked here shows no detected change either, once corrected for the number of variables tested at once (the raw trend clears IPCC's flat 10% example threshold but not the multiple-comparisons-adjusted bar -- see the Methodology page). |
| Asia: ACE (NWPAC+NIO) | computing… | Partially consistent AR6 reports an observed increase in the intensity of strong tropical cyclones affecting East and Southeast Asia, while the continent-wide ACE index here (NWPAC+NIO combined) shows no detected change -- plausibly diluted across two very different basins, not a contradiction of the more localized finding. |
IPCC AR6 Chapter 12 Regional Consistency — Europe
The cleanest match of any continent: AR6 reports an observed, continent-wide increase in the frequency and intensity of hot extremes for Europe (high confidence), directly matching the detected heat wave increase here since 1949, the only continent with a dedicated heat metric. Flooding also aligns well: AR6 finds sub-regional divergence (increasing in northwestern Europe, more uncertain elsewhere), the same pattern the cited literature (Blöschl et al. 2019) and the live index here (no detected continent-wide change) both show. Europe's wildfire finding shows a clean detected decrease on both tracked metrics, with no observed AR6 fire-weather statement (only a projected one) to compare it against.
Source: AR6 WGI Regional Fact Sheet — Europe — this site detects observed historical trends only, never projections, so this comparison draws only on AR6's observed findings, excluding its projected ones.
| Variable | GWX finding | AR6 Ch.12 consistency |
|---|---|---|
| Europe: % of land in D2+ (severe or worse) drought | computing… | Partially consistent AR6 reports an observed increase in hydrological, agricultural, and ecological drought specifically in the Mediterranean sub-region, while the continent-wide index here shows no detected change -- plausibly diluted by averaging a strong southern signal against the rest of Europe. |
| Europe: % of pour-point cells in high flow | computing… | Partially consistent AR6 reports an observed increasing trend in river flooding (Western & Central Europe) and an observed increase in pluvial flooding attributed to human influence (Northern Europe), while the continent-wide GloFAS index here shows no detected change -- consistent with a sub-regional signal diluting at continent scale. |
| Europe: dry matter combusted | computing… | Not directly comparable AR6's Europe fact sheet has no observed (only projected) fire-weather finding, and the wildfire records here are currently under the 30-year minimum record length in any case. |
| Europe: burned area | computing… | Not directly comparable AR6's Europe fact sheet has no observed (only projected) fire-weather finding, and the wildfire records here are currently under the 30-year minimum record length in any case. |
| Europe: most extreme storm's minimum pressure (annual) | computing… | Not directly comparable AR6's only storm-related finding for Europe (Northern Europe's increase in severe wind storms) is a projection, not an observed trend, and this site shows no detected change in storm intensity either way. |
| Europe: heat wave index (GSN stations) | computing… | Consistent AR6 reports an observed, continent-wide, high-confidence increase in the frequency and intensity of hot extremes for Europe, directly matching the detected heat wave increase here since 1949 on both tracked metrics. |
| Europe: heat wave magnitude (HWMId-style, GSN stations) | computing… | Consistent AR6 reports an observed, continent-wide, high-confidence increase in the frequency and intensity of hot extremes for Europe, directly matching the detected heat wave increase here since 1949 on both tracked metrics. |
IPCC AR6 Chapter 12 Regional Consistency — North America
AR6 reports observed increases in drought and fire weather across western and central North America, and observed increases in extreme precipitation in the east — but the continent-wide indices tracked here for drought, flooding, and both wildfire metrics currently show no detected change. This divergence most likely traces to AR6's sub-regional signals diluting when averaged across the whole continent (including Canada, Mexico, and Central America), rather than to any error in the underlying regional trends AR6 cites.
Source: AR6 WGI Regional Fact Sheet — North and Central America — this site detects observed historical trends only, never projections, so this comparison draws only on AR6's observed findings, excluding its projected ones.
| Variable | GWX finding | AR6 Ch.12 consistency |
|---|---|---|
| North America: % of land in D2+ (severe or worse) drought | computing… | Partially consistent AR6 explicitly reports observed drought increases in Western/Central North America and in Northern/Southern Central America and the Caribbean, while the continent-wide index here shows no detected change -- plausibly diluted across a very large, climatically diverse continent spanning Canada through Central America. |
| North America: % of pour-point cells in high flow | computing… | Not directly comparable AR6's North America fact sheet only projects future increases in river and pluvial flooding; it has no observed flood finding to compare the no-detected-change result here against. |
| North America: dry matter combusted | computing… | Not directly comparable AR6 reports an observed fire-weather increase in Western/Central North America, but fire-weather conditions (atmospheric favorability) and the realized-combustion/burned-area metrics tracked here measure different quantities, the same distinction drawn on the global Wildfire page; both wildfire records here also currently sit under the 30-year minimum record length regardless. |
| North America: burned area | computing… | Not directly comparable AR6 reports an observed fire-weather increase in Western/Central North America, but fire-weather conditions (atmospheric favorability) and the realized-combustion/burned-area metrics tracked here measure different quantities, the same distinction drawn on the global Wildfire page; both wildfire records here also currently sit under the 30-year minimum record length regardless. |
| North America: most extreme storm's minimum pressure (annual) | computing… | Not directly comparable AR6's only storm-related finding for North America (more extreme tropical cyclones, severe storms, and dust storms) is a projection, not an observed trend, and this site shows no detected change in extratropical storm intensity either way. |
| North America: heat wave index (US WSDI proxy) | computing… | Consistent North America's heat wave index defaults to the sibling US dashboard's WSDI headline metric (NOAA nClimGrid-Daily, 1951-present, added 2026-08-19) as a longer-record stand-in for the continent -- the US dominates North America's population and data density, and this record shows a detected increase (p=0.002), matching AR6's observed, human-attributed increase in extreme heat for the region. The continent-wide GSN station index (2000-present) remains selectable but still sits under the 30-year minimum on its own. |
| North America: heat wave magnitude (HWMId-style, GSN stations) | computing… | Consistent North America's heat wave index defaults to the sibling US dashboard's WSDI headline metric (NOAA nClimGrid-Daily, 1951-present, added 2026-08-19) as a longer-record stand-in for the continent -- the US dominates North America's population and data density, and this record shows a detected increase (p=0.002), matching AR6's observed, human-attributed increase in extreme heat for the region. The continent-wide GSN station index (2000-present) remains selectable but still sits under the 30-year minimum on its own. |
| North America: ACE (NATL+NEPAC) | computing… | Not directly comparable AR6's only tropical-cyclone finding for North America (more extreme storms) is a projection, not an observed trend, and this site shows no detected change in basin-combined ACE either way. |
IPCC AR6 Chapter 12 Regional Consistency — South America
AR6 reports high-confidence increases in drought frequency and duration in northern South America, matching the detected drought increase here for the continent. Flooding is the one place the numbers here disagree with each other, beyond disagreeing with AR6: the live GloFAS index shows a detected decrease in high-flow conditions, while separately cited literature (Barichivich et al. 2018) documents severe Amazon flooding roughly five times more frequent since the early 2000s — a tension between a broad continent-wide index and a well-documented regional case, left open rather than resolved by picking one.
Source: AR6 WGI Regional Fact Sheet — Central and South America — this site detects observed historical trends only, never projections, so this comparison draws only on AR6's observed findings, excluding its projected ones.
| Variable | GWX finding | AR6 Ch.12 consistency |
|---|---|---|
| South America: % of land in D2+ (severe or worse) drought | computing… | Consistent AR6 reports high confidence in a dominant increase in drought duration and frequency for northern South America, matching the detected drought increase here for the continent. |
| South America: % of pour-point cells in high flow | computing… | Not directly comparable AR6's Central and South America fact sheet has no observed continent-scale flood finding (its river/pluvial flood statements are projections, apart from one sub-region's observed increase in extreme precipitation, a different quantity); the live GloFAS index here also directly disagrees with a separate, independently cited literature finding (Barichivich et al. 2018) of a roughly fivefold increase in severe Amazon flooding since the early 2000s -- an internal tension, left open rather than resolved by picking one. |
| South America: dry matter combusted | computing… | Not directly comparable AR6's only fire-weather finding for this region (Southern Central America) is a projection, not an observed trend, and the wildfire records here are currently under the 30-year minimum record length regardless. |
| South America: burned area | computing… | Not directly comparable AR6's only fire-weather finding for this region (Southern Central America) is a projection, not an observed trend, and the wildfire records here are currently under the 30-year minimum record length regardless. |
IPCC AR6 Chapter 12 Regional Consistency — Oceania
AR6 reports opposing observed drought trends within Australia itself (decreasing in the north, increasing in the south) — plausibly why the continent-wide index here shows no detected change, since regionally opposite signals can net out in an aggregate. AR6 also reports an observed, continent-wide increase in fire weather days since 1950, while the burned-area and dry-matter-combusted metrics tracked here both show no detected change for Oceania — the same fire-weather-vs-realized-fire distinction drawn on the Wildfire page, not a contradiction.
Source: AR6 WGI Regional Fact Sheet — Australasia — this site detects observed historical trends only, never projections, so this comparison draws only on AR6's observed findings, excluding its projected ones.
| Variable | GWX finding | AR6 Ch.12 consistency |
|---|---|---|
| Oceania: % of land in D2+ (severe or worse) drought | computing… | Consistent AR6 reports opposing observed drought trends within Australia itself (a decrease in the north, an increase in the south, the southwest specifically attributed to human influence) -- exactly the kind of regionally opposite signal that plausibly nets out to the no-detected-change result here for the continent as a whole. |
| Oceania: % of pour-point cells in high flow | computing… | Not directly comparable AR6's Australasia fact sheet only projects future increases in heavy rainfall and river flooding; it has no observed flood finding to compare the no-detected-change result here against. |
| Oceania: dry matter combusted | computing… | Not directly comparable AR6 reports an observed, continent-wide increase in extreme fire weather days and a longer fire season since 1950, but fire-weather conditions and the realized-combustion/burned-area metrics here measure different quantities, the same distinction drawn on the global Wildfire page; both wildfire records here also currently sit under the 30-year minimum record length regardless. |
| Oceania: burned area | computing… | Not directly comparable AR6 reports an observed, continent-wide increase in extreme fire weather days and a longer fire season since 1950, but fire-weather conditions and the realized-combustion/burned-area metrics here measure different quantities, the same distinction drawn on the global Wildfire page; both wildfire records here also currently sit under the 30-year minimum record length regardless. |
| Oceania: ACE (SPAC+SIO) | computing… | Partially consistent AR6 reports an observed decrease in tropical cyclones for Northern Australia specifically -- directionally compatible with, though not identical to, the no-detected-change result here for the SPAC+SIO combined basin index. |