A slope change at four TSA degree heating weeks was uncertain in four Atlantic and Pacific coral samples in 2003-2019

We tested whether the heat slope of mean reported coral bleaching changed above four TSA degree heating weeks. Four separate models used 5,336 repeated-site monthly records from Atlantic and Pacific Reef Check surveys in 2003-2010 and 2011-2019. Models controlled for site, year, month and depth. High-dose minus low-dose slopes ranged from -16.36 to +1.75 percentage points per four TSA DHW. All four country-cluster intervals included zero after adjustment for four tests. Only 19 and 8 Atlantic months were above four. Two secondary negative changes did not repeat across all samples. This does not show a linear biological response or no heat effect. Sparse high-dose records, selected visits, missing reports, shared heat events and surface-dose error limit inference. The package includes source hashes, selected records, code, estimates, numerical checks and a figure. It tests a fixed curve shape, not an official alert threshold, air pollution or biodiversity loss.

coral bleaching · marine heat stress · nonlinear response · observation bias · regional replication · null result