Project Topics
Get Complete Project Material File(s) Now! »
Eddy covariance method
Based on wide acceptability, the EC method has stood out from other methods. It simply estimates fluxes as the covariance between the turbulent fluctuation of the vertical wind velocity ( ) and the mixing ratio ( ) (Equation 1.7). ̅̅̅̅̅̅′ (1.7) = ̅ ′0.
where is the dry air density, ′ is the fluctuation, the bar represents the mean, and the gas mixing ratio can be temperature for , water vapor for , or 2 density. From the early experimental campaigns (of Elagina et al., 1973; Elagina et al., 1978) that compared EC measurements with Bowen-ratio measurements, it became apparent that the energy balance at the Earth’s surface cannot be experimentally closed, and this is termed the ‘non closure problem’- where the sum of the turbulent fluxes ( + ) is less than the available energy ( − ). Later validation attempts of Equation 1.2 were made in Australia, 1981 (Leuning et al., 1982), in Germany, 1984 (Koitzsch et al., 1988), in Canada, 198(2-4) (Desjardins, 1985), in Russia, 1988 (Tsvang et al., 1991), in Estonia, 1990 (Foken et al., 1993), in USA, 1989 (Kanemasu et al., 1992), in Austria, 1989 (Bernhofer, 1992), in South Africa, 2000-2014 (Majozi et al., 2017), in China (Li et al., 2005), in (Wilson et al., 2002), in Germany (Imukova et al., 2016) etc. From these SEB studies, it is evident that irrespective of a site’s climate, location, instrumentation, vegetation type, phenology, etc), the surface energy can only be closed within a certain margin.
Modeling of the land surface budgets and fluxes
Like all in-situ systems, the aforementioned systems are spatially limited, cumbersome, and expensive. Mathematical models are alternative tools, and over the past decades, the flux community has experienced a surge of state-of-the-art models of various complexities. The following section presents some models for the simulation of and other SEB components.
Surface energy balance models
These models are grounded in the theory of the energy balance which requires a combination of ground-based and land surface temperatures ( ) to estimate by partially or fully solving the energy balance as a residual of the available energy from short-wave and long-wave radiation (Equation 1.2). In a broad sense, based on the parameterization of energy sinks and sources at the land-surface interface, these models can be divided into: single-source, dual-source, and multiple-source models (not discussed) (see Figure 1.4). For a proper structural representation of the surface energy transfer, a choice between these three is required.
Biophysical measurements
Over the studied years, we used two methods to measure the , and the Green Area Index ( ) which concerns the whole green matter. is measured using the destructive method. During the growing season, is monitored about five to six times. The sampling protocol involves collecting vegetation at 10 to 20 sampling points inside a footprint area representative of the crop plot. After vegetation collection, the leaves are separated from the stalks, then, these detached leaves are placed on the transparent bed of the Li-3100 planimeter (LiCor, Lincoln, NE, USA), and they are conveyed across the scanning bed that rapidly digitizes the area, length, and width of these leaves. Using these destructively-obtained measurements, a continuous dataset is constructed at a 30-mins time step using spline interpolation. Simultaneously, at these same sampling locations, measurements of the vegetation heights were obtained. Readers are referred to Béziat et al. (2009) for the detailed protocols for the different crops.
Since 2019, in accordance with the related ICOS protocol, a ceptometer (SS1 Sunscan Canopy Analysis System; Delta-T Devices), replaced this traditional method of estimation. For comparison purposes, in addition to this indirect method, the ICOS protocol requires a simultaneous acquisition of a destructively-obtained leaf area index particularly towards each crop’s vegetation peak. This is recommended because the Sunscan plant canopy analyzer system has the reputation to strongly underestimate the leaf area index due to its sensitivity to the photosynthetic active radiation under non-ideal light conditions (see Pokovai et al., 2019; Casa et al., 2019; Wilhem et al., 2000 for reviews).
Considering the fact that destructively-obtained ( in this section) serve as a reference measurement used in the validation of measurements from indirect methods, measurements taken between the 19th of June and 15th of July 2019 were used to correct the measurements from the Sunscan ( ). A very simple approach was adopted; hence, its interpretation should be made with absolute care because only 8 data points were available. Describing the relationship between and by a linear equation, Figure 2.3a presents a scatterplot comparing both raw measurements, and Figure 2.3b displays the outcome of the correction exercise. These new values ( ) would be used in chapter 5 in the estimation of the carbon fluxes for the year 2019.
Eddy covariance measurements
An eddy covariance system uses the eddy covariance method to measure exchanges between the atmosphere, and the underlying land surface. This method was proposed by Montgomery (1948), Swinbank (1951), and Obukhov (1951). The EC system samples eddies for their vertical velocity, and the concentration of the scalar of interest ( 2 , 2, 4, 2 etc). Figure 2.4 shows the flux towers installed at our study sites, and this offers the possibility of continuous flux measurement (see Table 1.1 in Foken et al. (2012) for the evolution of the EC system).
Figure 2.4: The eddy covariance flux tower installed at FR-Aur and FR-Lam with some components.
Turbulent exchanges at the surface (e.g., the sensible heat fluxes ( ), the latent heat fluxes ( ), the vertical turbulent 2 flux ( ) etc.) are measured by the EC system. An EC system consists of an Infra-Red Gas Analyzer (IRGA), and a three-dimensional sonic anemometer (HS50, Gill Ltd or CSAT, Campbell). The sonic anemometer measures the 3-D high frequency wind speeds and sonic temperature, while the IRGA measures the fluctuations in the air temperature, water vapor density, 2 density etc.
There are two types of IRGAs: the closed-path (CP), and the open-path (OP). In the CP set up, the concentration of the sampling air is quantified by drawing the air around the sonic through a sampling tube to the IRGA; whereas in the OP design, the IRGA is situated a few centimeters away from the anemometer, and the sampling air must move freely between the two systems (see Haslwanter et al. (2009) and Burns et al. (2015) for the pros and cons of each design). At our experimental sites, measurements running from 2005 to 2018 were made with an OP system (Li7500). Although, between 2013 and 2017, both OP and CP were run in parallel at the two sites. In 2018 and afterwards (at FR-Lam), only the CP system (Li7200) was retained in compliance with ICOS’s recommendation; and with this CP design, data loss due to precipitation is alleviated. In the former set-up, raw turbulent measurements were recorded at 20 Hz on a CR3000 data logger (Campbell Scientific Inc., Logan, UT, USA), and since 2018, they have been recorded at 10 Hz with a smart-flux data logger to ensure data synchronization between the sonic anemometer and the IRGA.
LANDSAT Land Surface Temperature
Launched in 1999 (2013), Landsat 7 (8) satellites continuously provided multi-spectral imageries of the Earth with one and two thermal bands, respectively. From 2005 to 2015, Land Surface Temperature ( ) maps (138 for FR-Lam) and (134 for FR-Aur) were retrieved from Landsat 7 (ETM+) and Landsat 8 (OLI & TIRS) sensors at a high spatial resolution of 30 m at approximately 10h 30, which coincides with the passing time of the satellite over FR-Lam and FR-Aur (both experimental sites were captured on the same scene). The retrieved thermal data were processed by the LANDARTs tool, and the report of this processing technique has been well documented in Tardy et al. (2016). The acquired maps were filtered by discarding scenes taken under cloudy conditions, and also erroneous scenes with missing data due to the mechanical failure of the Scan Line Corrector (SLC) in Landsat 7 since 2003. From the 71 scenes retained for FR-Lam and the 52 for FR-Aur, data were retrieved as geo-located digital values and then corrected for atmospheric and surface emissivity effects using the LANDARTs tool. These retrieved temperature values were used in chapter 3 to estimate the spatial and temporal variability of the surface temperature over FR-Lam and FR-Aur by adopting the methodology developed in Cuxart et al. (2016).
Eddy covariance data treatment
Time series, high-frequency EC data were post-processed with the EdiRe software package following the guidelines described in Lee et al. (2005), and Burba and Anderson (2010). Before the co-variances were calculated, outliers of six standard deviations from the population mean were removed from the time series. If four or more consecutive data points were detected with values larger than the standard deviation, then they were not considered as an extremity (Vickers and Mahrt, 1997). The time delay between the CSAT3 and Li-7500 was removed using a cross-correlation analysis. To ensure the CSAT3 is perfectly leveled, such that the vertical component ( ) is perpendicular to the mean streamline plane, the coordinates were rotated using the double rotation method (Aubinet et al., 2000). According to Lee et al. (2005), this method is suitable for ideal sites with little slope and fair-weather conditions.
The effects of density fluctuations induced by heat fluxes on water vapor measurements when using the Li-7500 were corrected using the procedure outlined by Webb et al. (1980). The spectral loss in the high frequency band due to path-length averaging, sensor separation, and signal processing was also corrected using Moore (1986). For the calculation of the sensible heat flux, the sonic temperature was converted to actual air temperature following the method of Schotanus et al. (1983).
Table of contents :
Chapter 1. State of the art on land-atmosphere exchanges
1.1 Concepts description
1.1.1 The surface energy balance
1.1.2 The water budget
1.1.3 The carbon budget
1.2 Observation techniques of land surface exchanges
1.2.1 Eddy covariance method
1.2.2 Scintillometry technique
1.2.3 Lysimetry
1.2.4 Sap flow method
1.3 Modeling of the land surface budgets and fluxes
1.3.1 Surface energy balance models
1.3.2 SVAT models
1.4 Carbon balance
Some main ideas of this chapter
Chapter 2. Site description, data, and model presentation
2.1 General introduction
2.2 Site description and data
2.2.1 Site description
2.2.2 Experimental data sets
2.2.2.1 Biophysical measurements
2.2.2.2 Meteorological measurements
2.2.2.3 Soil measurements
2.2.2.4 Eddy covariance measurements
2.2.2.5 Soil respiration measurements
2.2.2.6 Sap flow measurements
2.2.2.7 LANDSAT land surface temperature
2.2.2.8 MODIS leaf area index
2.2.3 Data processing
2.2.3.1 Eddy covariance data treatment
2.2.3.2 Partitioning of the net fluxes
2.3 Model description and implementation
2.3.1 ISBA-A-gs
2.3.1.1 The surface energy balance
2.3.1.2 Vertical transfer of water and energy within the soil
2.3.1.3 Description of the carbon processes
2.3.2 ISBA-A-gs-MEB
2.3.2.1 The surface energy balance
2.3.2.2 Additional parameterization
2.4 Forcing information
2.5 Model implementation
2.6 Definition of statistical metrics
Some main ideas of this chapter
Some general definitions
Chapter 3. Surface energy balance and flux partitioning of annual crops in southwestern France
3.1 Introduction
3.2 Methodology
3.2.1 Surface state categorization
3.2.2 Selection of contrasting years
3.2.3 The sensible heat advection term
3.3 Results
3.3.1 General considerations on SEB for both sites
3.3.1.1 Analysis of the ground heat flux term
3.3.1.2 How significant are the storage terms in the SEB?.
3.3.1.3 Annual and monthly variability of the energy balance closure
3.3.1.4 Dependence of the energy budget components on time of the day
3.3.2 Overview of the energy balance closure for contrasted years
3.3.2.1 Dependency on atmospheric parameters
3.3.3 Dependency of SEB on crop phenological stages and rainfall
3.3.3.1 Effect of the plant functioning on the SEB and its partitioning
3.3.3.2 Effect of rainfall
3.3.4 Impact of sensible heat advection on the surface energy balance
3.3.5 Effect of time averaging on the energy closure
3.4 Conclusions
Some main ideas of this chapter
Chapter 4. Estimation and partitioning of surface energy fluxes over a maize and wheat using a land surface model
4.1 Introduction
4.2 Materials and methods
4.3 Results and Discussion
4.3.1 Experimental data analysis
4.3.1.1 Meteorological conditions and vegetation characteristics
4.3.1.2 Energy balance closure..
4.3.2 Assessment of the ISBA and ISBA-MEB models
4.3.2.1 The Energy Budget…
4.3.2.2 The soil moisture
4.3.2.3 Transpiration
4.3.3 Inter-annual variability of the water budget
4.4 Conclusions
Some main ideas of this chapter
Chapter 5 Estimating carbon components over maize and wheat using a LSM
5.1 Introduction
5.2 Methodology
5.2.1 Sensitivity analysis and calibration
5.3 Results
5.3.1 Experimental data analysis
5.3.1.1 Comparison between three methods of soil respiration monitoring.
5.3.2 Assessment of ISBA-MEB for carbon exchanges
5.3.3 Net annual ecosystem production
5.4 Conclusions
Some main ideas of this chapter
Chapter 6 General conclusions (in English)
6.1 General conclusion
6.1.1 The surface energy balance closure
6.1.2 The energy and water budget
6.1.2 The carbon balance
Chapter 6 Conclusion générale (in French)
6.1 Conclusion générale
6.1.1 Le bilan énergétique de surface
6.1.2 Le bilan énergétique et hydrique
6.1.3 Le bilan carbone
Reference
