Print Click here to print this project

DEHESHyrE: Monitoring mass and energy fluxes in a manipulated Mediterranean tree-grass "Dehesa" ecosystem through the integration of ground and satellite data with airborne hyperspectral imagery

Start date: 01-04-2015 - End date: 31-07-2015

Status: Confirmed

Open to sharing: Yes

Confidential: No

Transnational Access: Yes

Open to training: No

Grounded / Maintenance: No


Aircraft name: CASA 212 RS - INTA

Airport: The study site is located in Las Majadas del Tietar, in the Province of Cáceres, where CEAM is operating a flux tower since May 2003. The site coordinates are longitude 5° 46’24.1’’ W and latitude 39° 56’29.4’’ N, at an approximate altitude of 260 m. The mean annual temperature is 16.7 ºC and the mean annual rainfall is about 650 mm. The ecosystem is a Holm oak open forest/savanna, with an understory dominated by herbaceous species and a few shrub types, named “dehesa” in Spain. Natural regeneration is good and the management consist continuous grazing. Tree density is approximately 20-25 per hectare, with a mean height of 8 m and a mean DBH of 405 mm. Canopy fraction cover is about 20%. The selected site offer ideal characteristics as it is both flat and horizontally homogeneous. Both characteristics are necessary for high quality eddy covariance flux measurements and very convenient for remote sensing analysis. The large extent of the homogeneous area around the flux-towers (> 1 km in all directions) ensures a good homogeneity within the flight overpass and the satellite sampled unit area (pixel size). Therefore, the potential bias due to differences in the areas which are sampled will be limited. This is an important factor when comparing ground-airborne and satellite products with flux-tower measurements. The two vegetation strata in the dehesa ecosystem (i.e. tree canopy layer and grassland) have different phenology and flux seasonal patterns which constitute an exciting spatial scaling challenge for DEHESHyrE.

Project description

Project theme: Integration of multitemporal/multiscale optical and thermal data to interpret and monitor ecosystem-scale water, carbon and nutrient fluxes in Mediterranean areas with complex vegetation structure

Science context: An intensive global effort has been put into modeling carbon and water exchanges between the terrestrial biosphere and the atmosphere. Keenan et al (2013) have stressed the importance of analyzing the ratio of water loss to carbon gain, or water-use efficiency, as this is a key characteristic of ecosystem function. Recent studies have shown the high potential of integrating ground, airborne-based and satellite observations to understand both the processes and the spatial scales governing the water and carbon cycles. However this integration remains a challenge, especially in ecosystems with complex vegetation structure as the "Dehesa"; a historical managed and integrated Mediterranean agro-forestry system which occupies large extensions in the Iberian Peninsula. It belongs to the class of tree-grass systems, which represent at least a third of the terrestrial land-surface. Despite their wide distribution, Earth observation systems and associated modeling have been so far poorly adapted to the key structural and functional characteristics of those systems (Pitman 2003). As a consequence, a significant uncertainty and bias in assessments of energy, carbon, water and biogeochemical dynamics has been often observed (Beringer et al. 2011). The objective of the proposed study is to better understand the contribution to the fluxes of both tree and grass layers, typical of these ecosystems, and the impact of large scale fertilization on biophysical and functional properties. Due to the different phenology of tree and grass components, this objective will be achieved by acquiring high spatial resolution airborne hyperspectral optical and thermal data during two different seasons, spring and summer.

Measurements to be made by aircraft: Scientific objectives DEHESHyrE proposes joint modeling of water, carbon and nutrient cycling by means of ground measurements, airborne and satellite systems, which involves combination of data at various spatial and temporal scales. Multiple efforts have been made to compare satellite data to ground-based biophysical parameters (Liang et al., 2011) because resolving the gap between different temporal and spatial scales is necessary. However, direct comparison between ground measurements and the spatial resolution of satellite data should be performed carefully and even avoided (Hufkens et al., 2008), unless structured sampling schemes are performed to facilitate multiple stage up-scaling processes (Anderson et al., 2004). Moreover, in highly variable landscapes, such as the one under investigation, comparing satellite data to ground observations may pose many problems and errors from averaging, aggregation, etc. (Fisher and Mustard, 2007). Airborne hyperspectral sensors will provide the required high spectral resolution data from the vegetation spectral response at a critical intermediate spatial scale between spaceborne and field sensors. The current proposal will contribute to this area of research considering: a) Spatial issues: related to the restricted spatial representativeness of carbon and water fluxes trough out the effective integration of optical and thermal sampling at Eddy Covariance (EC) sites with airborne and satellite data. The scale-appropriateness of the spectral measurements relative to the spatial sampling of EC will be investigated. b) Temporal issues: analyzing the dynamic of changes: daily (from continuous ground sensors) and seasonally (combining continuous ground sensor with high resolution airborne and satellite data). The overall objective of the experiment is to evaluate the potential of multi-source hyperspectral indices and thermal measurements to better understand the contribution of both tree and grass layers to the matter and water fluxes measured with the Eddy Covariance (EC) technique and to understand the impact of large scale fertilization on biophysical and functional properties of the study area. A specific objective will be the evaluation of the effect of different temporal and spatial sampling between EC measurements, ground, airborne and satellite optical and thermal data on water and carbon modeling at ecosystem scale. Proposed work The experiment will be conducted in a “dehesa” study site monitored with three EC systems. One EC tower has been operating in the study area since 2004 while 2 new systems will be installed from beginning of April 2014 in the context of a large scale nutrient manipulation experiment (MANIP) funded by Max Planck Institute. In 2015 the two towers footprint will be manipulated by fertilizing with N in one area and P in the other (or alternatively N and P together according to the outcome of an ongoing small scale experiment). The airborne survey, compatible with EUFAR schedule, should be planned in two different seasons (spring and summer) as the selected tree-grass ecosystem exhibit quite different structure/phenology and flux patterns. CASI 1500 and AHS-160 imagery will be acquired contemporary to intensive field campaigns. Field data, consisting in spectral measurements acquired with high resolution field spectroradiometers and biophysical vegetation parameters sampling, will be collected contemporary to the image acquisition in the footprint of the three EC towers (more details in the “plans for simultaneous field work plans” section). Airborne data will be related to ecosystem biophysical parameters and water and carbon fluxes using empirical approaches based on specific spectral bands and/or bands combinations (e.g. Photochemical Reflectance Index (PRI) and Normalized Difference Water Index (NDWI)) as well as physical models such as those based on radiative transfer theory. The availability of high resolution field spectroscopy data will allow also to test advanced techniques such as first derivative computation, band depth analysis, red-edge analysis, etc., as well as exploring new spectral indices and sun-induced chlorophyll fluorescence. The models developed with airborne data will be then validated against in situ measurements. Anticipated output Airborne data will be used to produce maps of ecophysiological and biophysical parameters, as well as canopy temperature, in the three footprint areas based on both semi-empirical and radiative transfer models. For example the photochemical reflectance index and Sun-Induced Fluorescence (SIF) will be used to derive information about ecosystem physiological response. These maps will be used to analyze if the changes in C/N/P stoichiometry induced by the manipulation experiment result in different ecosystem-physiological and structural response, as suggested by recent literature (Peñuelas et al. 2012). Furthermore with this experiment we expect to quantify the magnitude of the impact of N and P-availability to ecosystem-level CO2 uptake, water flux, and the resulting Water Use Efficiency (WUE). Anderson, M.C., Neale, C.M.U., Li, F., Norman, J.M., Kustas, W.P., Jayanthi, H., Chavez, J. 2004. Upscaling ground observations of vegetation water content, canopy height, and leaf area index during SMEX02 using aircraft and Landsat imagery, Remote Sensing of Environment, 92 (4), 447-464. Fisher, J.I. and Mustard, J.F., 2007. Cross-scalar satellite phenology from ground, Landsat, and MODIS data, Remote Sensing of Environment, 109 (3), 261-273. Hall, F.G., Hilker, T., Coops, N.C., Lyapustin, A., Huemmrich, K.F., Middleton, E., Margolis, H., Drolet, G., Black, T.A., 2008. Multi-angle remote sensing of forest light use efficiency by observing PRI variation with canopy shadow fraction, Remote Sensing of Environment, 112, 3201-3211. Hufkens, K., Bogaert, J., Dong, Q.H., Lu, L., Huang, C.L., Ma, M.G., Che, T., Li, X., Veroustraete, F., Ceulemans, R., 2008. Impacts and uncertainties of upscaling of remote-sensing data validation for a semi-arid woodland, Journal of Arid Environments, 72(8), 1490-1505. Liang, L.A., Schwartz, M.D., Fei, S.L., 2011. Validating satellite phenology through intensive ground observation and landscape scaling in a mixed seasonal forest. Remote Sensing of Environment, 115, 143–157. Peñuelas, J., Sardans, J., Rivas-ubach, A., Janssens, I.A., 2012. The human-induced imbalance between C, N and P in Earth's life system, Global Change Biology, 18, 3-6.

Season: First flight: spring 2015. Preferable during the second half of April-first half of May Second flight: summer 2015. Preferable during June or July. The two flights should be during the same growing season

Weather constraints: Clear sky is required in order to have homogeneous illumination conditions along the airborne scene. This would facilitate atmospheric correction as well as a higher chance to have cloud free satellite acquisitions in the same or close dates to the flights. In the selected acquisition area the likelihood of clear sky is important at any season and high from April to October. Statistics from the Spanish National Meteorological Agency reveals that in 2012 the province of Cáceres, in which the study area locates, had a total of 3192 sun hours per year with more than 300 hours per month from May to August.

Time constraints: Time of the day for the flight: around noon Passes of satellites: flight day simultaneous to Landsat 8 overpass would be desirable but not mandatory Weekends: Flight day during weekends would complicate the organization of simultaneous ground measurements due to restricted personnel and infrastructure availability Season: Two flights will be necessary to follow seasonal dynamics of the ecosystem, the first one during spring (April) and a second one during summer (June-July).

Flights (number and patterns): A temporal series of imagery would be the ideal solution to analyze spatial and temporal issues related to the scaling of carbon and water fluxes in the selected study area at ecosystem level, as the 2 layers (tree and grass), respond with different dynamics/phenology to the environmental factors. However, we assume that this is an unfeasible option and, therefore, we propose two AHS-CASI imagery campaigns in other to catch ecosystem dynamics coinciding with the most adequate dates according to the study site characteristics (vegetation growth cycle and illumination conditions), and the availability of simultaneous field and satellite data acquisition. The first flight would be planned in spring when the grass layer reach it biomass peak (usually in April) having an important contribution to water and carbon fluxes. A second flight in summer, when the grass layer is completely dry (June to September), would allow analyzing the contribution of the tree layer to the water and carbon fluxes. Each flight would acquire information around noon of a small area (1x1 km) around the three eddy-flux towers. The flight should have at least two swaths with a minimum 50% of overlap in order to have multi-angular observations needed to obtain more accurate retrievals of the vegetation structure as well as soil and vegetation temperatures. The along-track length area should not be very big (3-5km) in order that consecutive swaths are not very distant in time (<15min) and therefore the temperatures of consecutive scan lines could be comparable. We have estimated that, with the abovementioned requirements, the total flight required for acquisition of scientific data over the study area is probably less than one hour/flight campaign.

Instruments: None

Other constraints: No other constraints/requirements have been identified

Project website:

Scientific contact


PI email: