Mapping the productivity of Greek vegetation

Gross Primary Production (GPP) is the total amount of carbon that plants fix through photosynthesis. It is the largest carbon flux between the atmosphere and the land, and a key measure of ecosystem function, of crop and forest productivity, and of how vegetation responds to drought, heat and climate change.

Open the GPP dashboard

Light use efficiency models

GPP cannot be measured directly over large areas, so it is estimated with models. Light Use Efficiency (LUE) models are among the most widely used, as they are accurate yet relatively simple. They rest on one idea: plant productivity is proportional to the light the canopy absorbs.

GPP is therefore the product of absorbed photosynthetically active radiation (APAR) and the efficiency with which that energy is turned into fixed carbon (LUE). APAR can be derived from satellite observations of vegetation greenness. LUE changes with species, phenology and environmental conditions, so most models give each vegetation type a maximum efficiency (LUE_max) and reduce it with scalars that describe how unfavourable conditions limit photosynthesis.

The sCASE model

The GPP estimates are produced with sCASE, an LUE model originally developed as a near real-time GPP monitoring system for the forests of North Pindus National Park (Stagakis et al., 2015). Daily GPP is calculated as:

GPP = PAR × fAPAR × LUE_max × D_scalar × T_scalar × W_scalar × S_scalar

Satellite NDVI provides the fraction of absorbed radiation (fAPAR). The maximum efficiency and the temperature thresholds are assigned to each vegetation type from the vegetation map. LUE_max is then reduced by a series of scalars:

Flowchart of the sCASE model: MODIS gives NDVI and RDVI; ERA5 gives PAR, precipitation and temperature; the PFT map gives vegetation types. NDVI gives fAPAR, which with PAR gives APAR. RDVI gives Dscalar; precipitation and ETo give Wscalar; temperature gives Tscalar; Wscalar and PFT give Sscalar for sclerophylls. LUE equals LUEmax times the scalars, and GPP equals APAR times LUE. — Structure of the sCASE model as implemented for Greece. MODIS vegetation indices (NDVI, RDVI), ERA5-Land meteorological variables (PAR, precipitation, temperature) and the vegetation (PFT) map feed the calculation of absorbed radiation (APAR) and light use efficiency (LUE), whose product gives daily GPP. Click to enlarge.

Developmental scalar D_scalar: Derived from the RDVI vegetation index, it accounts for the stage of leaf development. Most productivity models lack it, yet it is particularly important for deciduous species.
Temperature scalar T_scalar: A bell-shaped response defined by the minimum, optimum and maximum temperatures for photosynthesis of each vegetation type.
Water scalar W_scalar: Based on the water balance between precipitation and reference evapotranspiration (ET₀, estimated from radiation and temperature), computed over a short (10-day) and a long (4-month) period, so that both brief dry spells and prolonged drought are captured.
Sclerophyll stress scalar S_scalar: An addition for the Greek implementation, applied only to sclerophyllous vegetation. It further reduces efficiency under severe water deficit, reflecting the strong drought response of Mediterranean evergreen shrublands.

sCASE was evaluated against eddy covariance GPP measurements from thirteen forest sites across Europe (Vanikiotis et al., 2026). Its daily estimates proved highly accurate at most sites, performing on par with other widely used LUE models.

GPP of Greece

sCASE runs in Google Earth Engine and produces daily GPP maps for the whole of Greece at 250 m resolution, from 2002 to the present. Vegetation indices come from the MODIS sensors on board the Terra and Aqua satellites, and the meteorological drivers from the ERA5-Land reanalysis.

A dedicated vegetation map of Greece assigns model parameters to each vegetation type, distinguishing the main forest genera (Abies, Pinus, Fagus, Quercus and others), sclerophyllous shrublands, olive groves, winter cereals and summer crops.

The interactive dashboard presents annual GPP maps, compares productivity among vegetation types, and shows the seasonal course and year-to-year variability of GPP together with its drivers: radiation, temperature, water availability and vegetation greenness.

Open the GPP dashboard

References

Stagakis, S., Markos, N., Vanikiotis, T., Tzotsos, A., Sykioti, O., Kyparissis, A. (2015). sCASE: A primary productivity monitoring system for the forests of North Pindus National Park (Epirus, Greece). European Journal of Remote Sensing, 48, 223–243. doi.org/10.5721/EuJRS20154813
Vanikiotis, T., Stagakis, S., Kyparissis, A. (2026). Evaluation of a LUE model and various water scalars based on eddy covariance data from 13 forest sites across Europe. Remote Sensing, 18(4), 548. doi.org/10.3390/rs18040548