Tab-contacts-off Tab-scientific-on


Additional research products (ARP4): NEFs (Non-Evaporative Fraction standardized)


Authors:

 

Mónica García, Juan Puigdefábregas (EEZA-CSIC)

 

Researchers at EEZA-CSIC have created a new product within the DeSurvey project related with the ecosystem water deficit, named NEFs (Non-Evaporative Fraction standardized)1.


It is targeted to assess and monitor, regionally and in spatially-explicit manner, land degradation processes occurring over natural ecosystems or dryland agriculture such as losses in soil depth, fertility and vegetation cover or density. Those processes should decrease the amount of water actually used by the ecosystem and increase the surface energy dissipated as sensible heat flux (H) with respect to a non-degraded status.

 

At a given time, NEFs represents the proportion of water that the ecosystem still has available to evapotranspire with respect to the proportion that would evapotranspire in a non-degraded situation. NEFs increases from 0 towards 1 with land degradation, e.g. NEFs=0.5 means that in that pixel and time the ecosystem is evapotranspiring 50% less water than it would do if was non-degraded.

 

The NEFs model is designed to run during a certain period of years at daily and annual time steps. Daily time series allow early detection of trends at altered sites, while annual time series provide more integrated and stable assessments. A spin-off product from NEFs is the Evaporation ratio (E/P) or the annual evapotranspiration divided by annual rainfall. It has been used to evaluate land degradation in the past 2. The NEFs model, besides retrieving degradation levels, quantifies the magnitude of changes in key variables of the surface energy balance and hydrologic cycle such as annual evapotranspiration or the sensible heat flux resulting from degradation providing additional information for land or watershed management.

 

Technical description

NEFs data requirements include remote sensing data in the thermal and optical ranges, a digital elevation model, maps of air temperature maps, and the aridity index (AI) map.

 

The model is implemented in three modules:

 

  1. Surface water deficit: estimates the daily surface water deficit as the ratio between H and net radiation or else using a modified version of the TVDI (Temperature Vegetation Dryness Index) 3 and potential evapotranspiration.
  2. Annual Integration. Daily/monthly evapotranspiration is calculated using water deficit estimates and potential evapotranspiration from A and it is subsequently integrated throughout the year obtaining annual evapotranspiration (E). E is divided by annual potential evapotranspiration obtaining the surface water deficit at annual scale. The E/P ratio can be estimated dividing E by annual precipitation.
  3. Standardizing. The surface water deficit from A or B is standardized in each climatic category of AI between a maximum and a minimum level of water deficit that is established spatially. Standardizing allows for regional applications of NEFs by comparing sites with different climatic conditions. Standardizing of the E/P allows using it for land degradation mapping as well.

 

Implementation

Different steps of the NEFs methodology have been tested using Aster (90 m), MODIS (1 km) and AVHRR (1 km) images at different time scales. The procedure to estimate the NEF and the intermediate variables involved in its calculation was validated both when using H/Rn and the TVDIt approach.

 

Comparisons with instrumental field observations in SE Spain provided acceptable error estimates similar to other models. The automatic procedure to find ecosystem extremes within each scene was successfully validated with field data. The impact of changing the spatial resolution of the model was evaluated using MODIS (1 km) and Aster (90 m).

 

Mean values were not significantly different when calculated with one sensor or another but regarding the scale, there was a lost of detail to be considered, although the dominant patterns are maintained. Field validation of NEFs in SE Spain at degraded vs. control field sites showed that NEFs detected significant differences between non-degrade sites and sites with reduced soil depth, associated with the loss of topsoil organic matter. Annual integration by estimating the E/P ratio was performed using the MEDOKADS database (NDVI and Ts at 1 km) between 1989-2000 for Andalusia (Spain) Annual spatial patterns for the E/P ratio in 95% of the study region were reasonable according to land use type and referenced values.

 

Notes

1 García M., Oyonarte C., Villagarcía, L., Contreras, S. Domingo, F. & Puigdefábregas, J. 2008b. Monitoring land degradation risk using ASTER data: the non-evaporative fraction as an indicator of ecosystem function. Remote Sensing of Environment, 112 (9): 3720-3736.


2 Boer, M.M. & Puigdefábregas, J., 2005: Assessment of dryland condition using spatial anomalies of vegetation index values. International Journal of Remote Sensing 26, (18): 4045-065


3 Sandholt, I. Rasmussen K. & Andersen, J. 2002: A simple interpretation of the surface temperature/vegetation index space for assessment of surface moisture status. Remote Sensing of Environment, 79: 213-224.