Although the biophysical effects of afforestation or deforestation on local climate are recognized, the biophysical consequences of seasonal and long-term dynamics in forests on understory microclimate, which creates microrefugia for forest organisms under global warming, remain less well understood. To fill this research gap, we combined a three-layered (i.e., canopy, forest air space and understory soil) land surface energy balance model and Intrinsic Biophysical Mechanism Model and quantify seasonal (warm minus cool seasons) and long-term changes (later minus former periods) in the biophysical effects of forest dynamics on understory air temperature (Δ
T
a
) and soil surface temperature (Δ
T
s
). We found that high latitudes forests show strongest negative seasonal variations in both Δ
T
a
and Δ
T
s
, followed by moderate latitudes forests. In contrast, low latitudes forests exhibit positive seasonal variations in Δ
T
a
and weak negative seasonal variations in Δ
T
s
. For the long-term variations, Δ
T
s
increases systematically at all three latitudes. However, the situation differs greatly for Δ
T
s
, with a weak increase at low and moderate latitudes, but a slight decrease at high latitudes. Overall, changes in sensible and latent heat fluxes induced by forest dynamics (such as leaf area index), by altering the aerodynamic resistances of canopy and soil surface layers, are the main factors driving changes in forest microclimate effects. In addition, this study also develops an aerodynamic resistance coefficient
f
r
1
to combine the air temperature effects and surface soil temperature effects and proposes an indicator – Δ
T
Su
, that is,
Δ
T
Su
=
Δ
T
s
+
(
1
f
r
1
−
1
)
Δ
T
a
, as a possible benchmark for evaluating the total biophysical effects of forests on temperatures.