?
Observing whole-canopy short-term dynamic response to natural step changes in incident light: Characteristics of tropical and temperate forests.
We examine the physical and biological responses of forest canopies to step changes in light
caused by passing low cumulus clouds that intermittently block the direct solar beam. Using
data obtained at a tropical rainforest and at a midlatitude deciduous forest, we estimate the
course of sensible heat flux, net ecosystem exchange, evapotranspiration, and water-use effi-
ciency in response to the rapid changes in the incident radiative flux. To describe these fluxes
during the interval over which the effects of stomatal time delays can be most influential,
eddy fluxes are estimated over minute or shorter intervals by invoking a conditional-sampling
procedure based on forming a Reynolds-average ensemble. The most important differences
between the two forests’ physical responses are in the thermal balances and heat-flux time
response constants. During the initial period after the light transition the only mean variables
that show appreciable changes are the blackbody and air temperatures, the other scalars being
little affected. We find that a distinct transient thermal internal boundary layer appears ≈ 20 m
thick above the temperate deciduous forest and ≈ 45 m thick above the tropical rainforest.
At each forest, the effective thickness of the inferred thermal outer-canopy ‘big leaf’ is about
1 mm. Twenty minutes after the abrupt change in incident light, ensemble eddy-flux esti-
mates approach those found using conventional time averaging, confirming the validity of the
ensemble approach. Previously unrecognized transient maxima in net ecosystem exchange
and evapotranspiration are evident 5–10 min following the shadow-to-light transition, longer
than the average light interval between shadows observed on partly-cloudy days in each case.
Short-term variations in sensible heat flux, net ecosystem exchange, and evapotranspiration
approximate an exponential adjustment, implying that first-order time-dependent single-leaf
models are adequate to describe whole-canopy processes in these cases, providing an experi-
mental method for determining whole-canopy bulk stomatal time constants. During the sunlit
interval (direct and diffuse radiative fluxes combined) net ecosystem exchange is enhanced,
while under cloud shadow (only diffuse radiative flux) water-use efficiency increases. This light and shadow alternation provides a mechanism describing the observed enhanced net
ecosystem exchange and water-use efficiency under certain types of partly-cloudy sky. We
apply empirical flux-response curves to an idealized case of radiative flux varying in a regular
on–off light and shadow pattern. For this case, an analytical solution for mean net ecosystem
exchange flux as a function of diffuse-radiation fraction yields results that strongly resemble
previous findings based on conventional time-averaged fluxes. Our analysis and modelling
indicates that a well-known correlation between diffuse radiative flux and enhanced net
ecosystem exchange and water-use efficiency on cloudy days is in many cases not causal,
but rather to be a consequence of time averaging over light-and-shadow intervals. By linking
processes associated with photosynthesis in fluctuating light at the leaf scale to the canopy
scale, our efforts facilitate the scaling-up of leaf responses to the ecosystem scale.