 |
| Titel |
Estimating gross primary productivity (GPP) of forests across southern England at high spatial and temporal resolution using the FLIGHT model |
| VerfasserIn |
Prasan Pankaew, Edward Milton, Terry Dawson, Jadu Dash |
| Konferenz |
EGU General Assembly 2013
|
| Medientyp |
Artikel
|
| Sprache |
Englisch
|
| Digitales Dokument |
PDF |
| Erschienen |
In: GRA - Volume 15 (2013) |
| Datensatznummer |
250078932
|
|
|
|
|
|
| Zusammenfassung |
| Forests and woodlands play an important role in CO2 flux and in the storage of carbon,
therefore it is important to be able to estimate gross primary productivity (GPP) and its
change over time. The MODIS GPP product (MOD17) provides near-global GPP, but at
relatively coarse spatial resolution (1km pixel size) and only every eight days. In
order to study the dynamics of GPP over shorter time periods and over smaller
areas it is necessary to make ground measurements or use a plant canopy model.
The most reliable ground-based GPP data are those from the FLUXNET network,
which comprises over 500 sites worldwide, each of which measures GPP using
the eddy covariance method. Each FLUXNET measurement corresponds to GPP
from an area around the sampling tower, the size and shape of which varies with
weather conditions, notably wind speed and direction. The FLIGHT forest light
simulation model (North, 1996) is a Monte Carlo based model to estimate the GPP
from forest canopies, which does not take into account the spatial complexity of
the site or the wind conditions at the time. Forests in southern England are small
and embedded in a matrix of other land cover types (agriculture, urban etc.), so
GPP estimated from FLIGHT needs to be adjusted to match that measured from a
FLUXNET tower. The aim of this paper is to develop and test a method to adjust
FLIGHT GPP so that it matches FLUXNET GPP. The advantage of this is that GPP
can then be estimated over many other forests which do not possess FLUXNET
sites.
The study was based on data from two mixed broadleaf forests in southern England
(Wytham Woods and Alice Holt forest), both of which have FLUXNET sites located
within them. The FLUXNET meteorological data were prepared for use in the
FLIGHT model by converting broadband irradiance to photosynthetically active
radiance (PAR) and estimating diffuse PAR, using methods developed in previous
work by the authors. The standard FLIGHT model tended to overestimate GPP
in the winter and spring period and under-estimate GPP in the summer months.
Correction factors were computed based on the midday GPP for each month of the
year.
The modified FLIGHT model was used to estimate GPP from each of the two forest
sites at hourly intervals over a year. Both sites showed a strong linear relationship
between GPP estimated from FLIGHT and GPP measured by FLUXNET (Alice
Holt forest, R2=0.96, RMSE = 2.39 μmol m-2 s-1, MBE = 1.32 μmol m-2 s-1 ,
Wytham Wood R2 = 0.97, RMSE = 1.42 μmol m-2 s-1, MBE = 0.57 μmol m-2
s-1).
The results suggest that the modified FLIGHT model could be used to estimate GPP at
hourly intervals over non-instrumented forest sites across southern England, and thereby
obtain regional estimates of GPP at high spatial and temporal resolution.
Reference
North, P. R. J. (1996). Three-Dimensional Forest Light Interaction Model Using a Monte
Carlo Method. IEEE Transactions on Geoscience and Remote Sensing, 34(4), 946-956. |
|
|
|
|
|
|