I read Antony Beevor cite the standard 40,000 civilian death toll from the week of air raids on Stalingrad in 1942. Beevor is a credible author, but could that death toll really have been so high? That figure is a little below Hamburg in 1943, and a little above Dresden in 1945 by most tallies. This makes Stalingrad the second most lethal air attack series in Europe of all of WWII.
Sure, the Soviets probably lacked quality bomb shelter networks. But the raids against Hamburg and Dresden were done by big four-engine US and British bombers. Stalingrad was bombed with two-engine JU-88's and HE-111's. That, and Stalingrad was no Hamburg population-wise. And you could say refugees added to the total, except Dresden also had refugees. Finally, 40,000 is a round enough number to be suspicious. Thanks for your thoughts!
For what it's worth, Richard Overy in The Bombing War: Europe, 1939-1945 estimates that the 40,000 figure is an overestimate, for the following reasons.
This estimate is supposed to come from a specific bombing raid (August 23, 1942) on industrial installations and communications networks, plus the city center the following day. The extensive damage was possible because of poor anti-aircraft defenses, meaning the Luftwaffe could fly at only 2,000-3,000 meters, and the fact that water mains were knocked out, meaning the fires could not properly be fought. Stalingrad was in fact bombed over six days (August 23 to August 29), by a collection of 400 Ju88 and He111 bombers, and then over the course of September, so the destruction wasn't just a single catastrophic day.
The prewar population of Stalingrad was some 440,000 civilians, many of whom were evacuated before the battle. Overy notes that an almost 10% death rate in a city attacked by a non-atomic bombing raid in World War II was otherwise unheard-of. Civil Defense authorities at the scene recorded some 3,931 deaths between July and October (mostly at the river crossings), and while these records are no doubt incomplete they are an order of magnitude less than the 40,000 figure.
The city was modern, with wide boulevards, parks and many concrete and stone buildings, so a force of 400 German aircraft dropping some 1,000 tons of explosives wouldn't be able to generate a firestorm capable of killing 40,000 people (as happened at Hamburg). 1,000 -2,000 deaths, however, would be in line with similar raids conducted by the Luftwaffe on London.
Nevertheless, even these reduced figures would make the victims of aerial bombing in Stalingrad the highest number in any bombed Soviet city during the war.
As for where the 40,000 figure originally comes from...that is a good question. Overy cites Joel Hayward's Stopped at Stalingrad, but otherwise notes this is a "standard" figure.