@prefix dcterms: . @prefix np: . @prefix npx: . @prefix orcid: . @prefix prov: . @prefix rdfs: . @prefix schema: . @prefix sub: . @prefix this: . @prefix xsd: . sub:Head { this: a np:Nanopublication; np:hasAssertion sub:assertion; np:hasProvenance sub:provenance; np:hasPublicationInfo sub:pubinfo . } sub:assertion { sub:schools a schema:CreativeWork; dcterms:creator orcid:0000-0002-1784-2920; rdfs:comment """Checking a climate result the honest way When scientists publish a result, others should be able to get the same answer with different tools. We tried exactly that with a marine-heatwave study — a good example of how science checks itself. The question — A 2018 paper reported that marine heatwaves — spells when the sea surface is unusually hot — became more frequent and longer. We re-tested one part: did the number of marine-heatwave days per year really rise between 1982 and 2016? The test — We deliberately swapped the ingredients. Instead of the American sea-surface-temperature data the original used, we used satellite data from the European Space Agency; instead of the original team's own program, we used a different, independent one. The rule for what counts as a “heatwave” was kept identical, so the comparison was fair. The result — The numbers matched closely: about 32 extra heatwave days over the period against the 30 the original found — within about 6%. Heatwave frequency and length also rose, agreeing with the original. And the trend stayed even after we removed the influence of El Niño. Being honest about limits — We were careful to say what we could NOT check: the paper's famous “54% over a century” figure needs daily records from before 1981 that don't exist independently, so we left it clearly out of scope. Good science states what it can and cannot show. The lesson — Getting the same answer with different data and different software is strong evidence, because a shared dataset or shared code could hide the same mistake in both — and here nothing was shared. That is what independent replication is for. Same answer from independent tools = strong evidence — and saying what you didn't test matters just as much."""; schema:about ; schema:audience . } sub:provenance { sub:assertion prov:generatedAtTime "2026-08-15T17:12:52.532557+00:00"^^xsd:dateTime; prov:wasAttributedTo . } sub:pubinfo { this: prov:generatedAtTime "2026-08-15T17:12:52.532557+00:00"^^xsd:dateTime; prov:wasAttributedTo . sub:sig npx:hasAlgorithm "RSA"; npx:hasPublicKey "MIIBIjANBgkqhkiG9w0BAQEFAAOCAQ8AMIIBCgKCAQEAyuQ7Hi/zQ7hDZtMtLWK1JKU83Kgo8NqC+b8sxD5IoiekPctXVA8iM0zD/9pcp1txGXjX/fUj5HRUkOn4RVHP3fFa/3irQYpZFQ5fHRj8MQ7kJ6AEd1oggfajDevZGz63hj6Vigo+BNYBXqhcbWVAfLfbzQPAegce9zH6n9vTqKYMrwrSxPQdhoAcoaKOqY8gRiJgzekeEr+iEqa63X1HwWtYDs7X25a9xWbgaDSsRjzZGVUVOn1vYdSfWAaIw+EB9CiJfKm1zuXcur5PDziMW1pnZG6/NQtrA7AM3gLnoSoU3glkm5UhksGNqu2SV+j8oZLg7v6C5/sg2f6LIJMeFwIDAQAB"; npx:hasSignature "eSn/3Lf6VdIIbotyJ1XdYhujWjnsLuyJxpfM0yAOTQ3ygaDVyElifPhPD7diViYpSAWv4eNoFus3KE6uWcSd4NEwoVdOqriw8tr/JDdd/CnmPX64vZWukHdBhPV2btfQadZXL6Pe+E7s0dX5CTRb6alHmEOREAETRqeESb7hEBB7YPtOXmSlVtOKrBu/vq7vxo/srH1EjFM7nuiCllMkdA4+XVqUXVAEv5+dTyQB+wIOMc/iJdavbNM17MlC72oELgPnRfKqD9ehAKAhdq+gjPcSTVPLormMjvxV9j1iXzqbEh6zOLUqxpkkKM4m5SQ+jhdrt2OTelBhVDbA0kGoag=="; npx:hasSignatureTarget this:; npx:signedBy . }