A full "current (2022) knowledge" review on the topic of tree dormancy triggers:
https://www.cell.com/current-biology/fulltext/S0960-9822(22)00580-2
One paragraph on environment:
Environmental signals controlling the annual growth cycle
How does the tree ‘know’ when to stop growing in the autumn and when to resume growth again in the spring? In many perennials, the dominant signal controlling the onset of growth cessation and bud set is day length (photoperiod). It is the shortening of the day length in late summer and early autumn that is the signal that triggers the plant to stop growth (
Figure 2). The rationale for this is thought to be that day length is a much more solid predictor of the time of the year than temperature, given that there can be large fluctuations in temperatures diurnally, seasonally and between years. However, in some trees, such as apple, pear, and mountain ash, a decrease in temperature is the most important signal. For trees responding to photoperiod, there is a ‘critical day length’ below which the tree will enter growth cessation (sometimes, and more physiologically correctly, also referred to as ‘critical night length’). The critical day length is central for the ability of trees to adapt to growth at different latitudes. The further north they grow, the longer the critical day length, because as one moves north, winter will arrive earlier, meaning that growth has to be stopped earlier in the season. However, an additional consideration is that trees growing at northern latitudes experience longer days during the summer compared with trees growing further south. This means that an aspen tree growing in southern Germany will stop growing when the days are shorter than 16 hours long, while an aspen tree growing in northern Sweden will already have ceased its growth when the day length drops below 21 hours (or even longer)! The critical day length threshold is genetically determined and, if the German tree is transplanted in northern Sweden, it will grow much longer into the autumn compared with its Swedish relative. However, it will then suffer, or even die, during the first winter because it will not have had enough time to develop winter hardiness. In contrast, the Swedish tree planted in Germany will hardly grow at all, setting bud prematurely soon after growth is resumed. This shows that the tree ‘chooses’ to stop growing at a particular time, not because the conditions are no longer permissible, but because it anticipates the arrival of winter at a certain time. There is, therefore, an important trade-off between adaptation to winter conditions and the length of the growing season.
(I have stubborn tridents myself that hold onto leaves well after the third or fourth hard frost) many vary slightly but the bottom line is pretty much as I described it from what I understand.
