Architects typically don't know such things (I went to architecture school), they hire engineers to compute the effect. This is typical practice: The architect

Architects typically don't know such things (I went to architecture school), they hire engineers to compute the effect. This is typical practice: The architect draws intent, then hires someone to compute effect, then there are meetings about the difference between intent and effect, and modifications, and a loop, and eventually the thing is built. That one person does intent and another computes effect is deliberate. A kind of four-eyes principle. Then more eyes look at it when the plans are submitted to the city's authority, but that isn't usually as thorough as the review cycle between architect and hired engineers. In this case they computed that the peak heat would be 36°C, the actual peak afterwards is given as 72°C. Those are very different temperatures. FWIW, based on my somewhat primitive knowledge of this (I didn't finish that school) the most likely cause for the difference is imperfect modelling of the wind. How buildings affect wind is still an active area of research. Both this building and its neighbour affect air circulation in ways that couldn't be modelled using the current state of the art. When focused sunlight works on calm air, it can raise the temperature a great deal. Warm air staying calm instead of rising is a surprise, of course. (The poor modelling is common BTW, although usually not that serious. In my native Norway snow drifts often collect in surprising spots around new buildings, because the building produces calm air here, wind there.)