Characterising groundwater use by vegetation using a surface energy balance model and satellite observations of land surface temperature

This study presents a novel ‘model-data’ approach to detect groundwater-dependent vegetation (GDV), through differences in modelled and observed land surface temperatures (LST) in space and time. Vegetation groundwater use is inferred where modelled LST exceeds observed LST by more than a threshold determined from consideration of systematic and random errors in model and observations. Modelled LST was derived from a surface energy balance model and LST observations were obtained from Terra-MODIS thermal imagery. The model-data approach, applied in the Condamine River Catchment, Queensland, Australia, identified GDV coincident to existing mapping. GDV were found to use groundwater up to 48% of the time and for as many as 56 consecutive days. Under driest of conditions, groundwater was estimated to contribute up to 0.2 mm h⁻¹ to total ET for GDV. The ability to both detect the location and water-use dynamics of GDV is a significant advancement on previous remote-sensing GDV methods.     

Modelling heat,water and carbon fluxes in mown grassland under multi-objective and multi-criteria constraints

A Monte Carlo-based calibration and uncertainty assessment was performed for heat, water and carbon (C) fluxes, simulated by a soil-plant-atmosphere system model (CoupModel), in mown grassland. Impact of different multi-objective and multi-criteria constraints was investigated on model performance and parameter behaviour. Good agreements between hourly modelled and measurement data were obtained for latent and sensible heat fluxes (R² = 0.61, ME = 0.48 MJ m⁻² day⁻¹), soil water contents (R² = 0.68, ME = 0.34%) and carbon-dioxide flux (R² = 0.60, ME = −0.18 g C m⁻² day⁻¹). Multi-objective and multi-criteria constraints were efficient in parameter conditioning, reducing simulation uncertainty and identifying critical parameters. Enforcing multi-constraints separately on heat, water and C processes resulted in the highest model improvement for that specific process, including some improvement too for other processes. Imposing multi-constraints on all groups of variables, associated with heat, water and C fluxes together, resulted in general effective parameters conditioning and model improvement.     

Multi-sensor model-data fusion for estimation of hydrologic and energy flux parameters

Model-data fusion offers considerable promise in remote sensing for improved state and parameter estimation particularly when applied to multi-sensor image products. This paper demonstrates the application of a ‘multiple constraints’ model-data fusion (MCMDF) scheme to integrating AMSR-E soil moisture content (SMC) and MODIS land surface temperature (LST) data products with a coupled biophysical model of surface moisture and energy budgets for savannas of northern Australia. The focus in this paper is on the methods, difficulties and error sources encountered in developing an MCMDF scheme and enhancements for future schemes. An important aspect of the MCMDF approach emphasized here is the identification of inconsistencies between model and data, and among data sets.The MCMDF scheme was able to identify that an inconsistency existed between AMSR-E SMC and LST data when combined with the coupled SEB-MRT model. For the example presented, an optimal fit to both remote sensing data sets together resulted in an 84% increase in predicted SMC and 0.06% increase for LST relative to the fit to each data set separately. That is the model predicted on average cooler LST's (∼ 1.7 K) and wetter SMC values (∼ 0.04 g cm^− 3) than the satellite image products. In this instance we found that the AMSR-E SMC data on their own were poor constraints on the model. Incorporating LST data via the MCMDF scheme ameliorated deficiencies in the SMC data and resulted in enhanced characterization of the land surface soil moisture and energy balance based on comparison with the MODIS evapotranspiration (ET) product of Mu et al. [Mu, Q., Heinsch, F.A, Zhao, M. and Running, S.W. (in press), Development of a global evapotranspiration algorithm based on MODIS and global meteorology data, Remote Sensing of Environment.]. Canopy conductance, g_C, and latent heat flux, λE, from the MODIS ET product were in good agreement with RMSEs for g_C = 0.5 mm s^− 1 and for λE = 18 W m^− 2, respectively. Differences were attributable to a greater canopy-to-air vapor pressure gradient in the MCMDF approach obtained from a more realistic partitioning of soil surface and canopy temperatures.     

Regional-scale analysis of carbon and water cycles on managed grassland systems

Predicting regional and global carbon (C) and water dynamics on grasslands has become of major interest, as grasslands are one of the most widespread vegetation types worldwide, providing a number of ecosystem services (such as forage production and C storage). The present study is a contribution to a regional-scale analysis of the C and water cycles on managed grasslands. The mechanistic biogeochemical model PaSim (Pasture Simulation model) was evaluated at 12 grassland sites in Europe. A new parameterization was obtained on a common set of eco-physiological parameters, which represented an improvement of previous parameterization schemes (essentially obtained via calibration at specific sites). We found that C and water fluxes estimated with the parameter set are in good agreement with observations. The model with the new parameters estimated that European grassland are a sink of C with 213 g C m⁻² yr⁻¹, which is close to the observed net ecosystem exchange (NEE) flux of the studied sites (185 g C m⁻² yr⁻¹ on average). The estimated yearly average gross primary productivity (GPP) and ecosystem respiration (RECO) for all of the study sites are 1220 and 1006 g C m⁻² yr⁻¹, respectively, in agreement with observed average GPP (1230 g C m⁻² yr⁻¹) and RECO (1046 g C m⁻² yr⁻¹). For both variables aggregated on a weekly basis, the root mean square error (RMSE) was ∼5–16 g C week⁻¹ across the study sites, while the goodness of fit (R²) was ∼0.4–0.9. For evapotranspiration (ET), the average value of simulated ET (415 mm yr⁻¹) for all sites and years is close to the average value of the observed ET (451 mm yr⁻¹) by flux towers (on a weekly basis, RMSE∼2–8 mm week⁻¹; R² = 0.3–0.9). However, further model development is needed to better represent soil water dynamics under dry conditions and soil temperature in winter. A quantification of the uncertainties introduced by spatially generalized parameter values in C and water exchange estimates is also necessary. In addition, some uncertainties in the input management data call for the need to improve the quality of the observational system.     

A distributed model of water balance in the Motueka catchment,New Zealand

A distributed water balance model is used to simulate the soil moisture regime of the Motueka catchment. The model is a major simplification of the Distributed Hydrology–Vegetation–Soil Model (DHVSM) with modifications suitable for the study area. The model was applied at 25-m resolution with a 1-day time-step for 10 years. The simulated hydrograph showed good correspondence with the observed hydrograph and there was good agreement of simulated and measured mean annual discharges (57.3 m³ s⁻¹ as compared with 58.7 m³ s⁻¹). Five different land cover scenarios were used to predict the effects of vegetation change on the hydrological regime: (1) current land cover; (2) prehistoric land cover; (3) maximum pine planting; (4) pine trees on easy slopes; and (5) pine trees on steep slopes. The pine scenarios all reduced the mean annual flow by about 2 m³ s⁻¹, while the prehistoric scenario reduced the mean annual flow by about 6 m³ s⁻¹. The pine scenarios (3, 4, and 5) reduced the 7-day 5-year low flow from 7.4 m³ s⁻¹ to between 6.5 m³ s⁻¹ and 6.8 m³ s⁻¹, respectively; and the prehistoric scenario reduced the 7-day 5-year low flow to 5.3 m³ s⁻¹.     

Influence of woody elements of a Norway spruce canopy on nadir reflectance simulated by the DART model at very high spatial resolution

A detailed sensitivity analysis investigating the effect of woody elements introduced into the Discrete Anisotropic Radiative Transfer (DART) model on the nadir bidirectional reflectance factor (BRF) for a simulated Norway spruce canopy was performed at a very high spatial resolution (modelling resolution 0.2 m, output pixel size 0.4 m). We used such a high resolution to be able to parameterize DART in an appropriate way and subsequently to gain detailed understanding of the influence of woody elements contributing to the radiative transfer within heterogeneous canopies. Three scenarios were studied by modelling the Norway spruce canopy as being composed of i) leaves, ii) leaves, trunks and first order branches, and finally iii) leaves, trunks, first order branches and small woody twigs simulated using mixed cells (i.e. cells approximated as composition of leaves and/or twigs turbid medium, and large woody constituents). The simulation of each scenario was performed for 10 different canopy closures (CC = 50-95%, in steps of 5%), 25 leaf area index (LAI = 3.0-15.0 m² m^− 2, in steps of 0.5 m² m^− 2), and in four spectral bands (centred at 559, 671, 727, and 783 nm, with a FWHM of 10 nm). The influence of woody elements was evaluated separately for both, sunlit and shaded parts of the simulated forest canopy, respectively. The DART results were verified by quantifying the simulated nadir BRF of each scenario with measured Airborne Imaging Spectroradiometer (AISA) Eagle data (pixel size of 0.4 m). These imaging spectrometer data were acquired over the same Norway spruce stand that was used to parameterise the DART model.The Norway spruce canopy modelled using the DART model consisted of foliage as well as foliage including robust woody constituents (i.e. trunks and branches). All results showed similar nadir BRF for the simulated wavelengths. The incorporation of small woody parts in DART caused the canopy reflectance to decrease about 4% in the near-infrared (NIR), 2% in the red edge (RE) and less than 1% in the green band. The canopy BRF of the red band increased by about 2%. Subsequently, the sensitivity on accounting for woody elements for two spectral vegetation indices, the normalized difference vegetation index (NDVI) and the angular vegetation index (AVI), was evaluated. Finally, we conclude on the importance of including woody elements in radiative transfer based approaches and discuss the applicability of the vegetation indices as well as the physically based inversion approaches to retrieve the forest canopy LAI at very high spatial resolution.     

Mobile phone use among college students is a sedentary leisure behavior which may interfere with exercise

Research suggests that cell phone use is related to sedentary behavior, that cell phone use during exercise reduces intensity, and that high frequency cell phone users are less fit than other users. Thus, cell phone use appears connected to health and fitness behaviors and should be better understood within this context. The present study investigated the sedentary nature of cell phone use, and examined the likelihood of cellphone use interfering with exercise behavior.DesignA validated survey was administered to a random sample of students from a public US university (N = 226).ResultsMean self-reported cell phone use was 380 min day⁻¹, 87% reported cell phone use primarily occurs while sitting, and 70% of use was for leisure. Cell phone use was positively related to sedentary behavior (β = 0.30, p < 0.001). It was not related to physical activity. However, the likelihood of cell phone use during moderate (p = 0.006) and mild (p < 0.001) intensity exercise increased as cell phone use increased.ConclusionLike other screens (e.g., TVs), cell phone use appears to be a sedentary leisure behavior. Furthermore, high frequency use increases the likelihood that it will occur during exercise, likely lowering exercise intensity.     

Predicting the thermodynamics of aluminum dross denitrification of flue gas

Existing alumina extraction and material production methods result in the formation of harmful ammonia gas or ammonia water originating from aluminum nitride (AlN) in dross. Therefore, in this study, aluminum dross was used as a denitration reagent to eliminate nitrogen oxides in flue gas and AlN in dross. Based on the proposed scheme, thermodynamic calculations were performed to investigate the denitrification effect and reduction of aluminum dross in flue gas. The results show that equilibrium concentrations of NO, NO₂, and HF in the flue gas were influenced mainly by temperature; their concentrations increased with an increase in the temperature, reaching 4.4 × 10⁻²⁰, 1.7 × 10⁻³⁸, and 7.0 × 10⁻⁸ g/m³, respectively, at 923 K. The Gibbs free energy corresponding to the reaction of CO₂ with Al/AlN in aluminum dross was −377/–120 kJ/mol. HF, originating from the reaction of NaF and water vapor, maintained an extremely low concentration of 6.99 × 10⁻⁸ g/m³ at 923 K. These results indicate that aluminum dross processing may clean the flue gas and increase the calorific value while eliminating the hazards of AlN. The results obtained herein will provide theoretical guidance toward new avenues of aluminum dross utilization.     

Mapping burn severity and burning efficiency in California using simulation models and Landsat imagery

Uncertainties in burning efficiency (BE) estimates can lead to large errors in fire emission quantification (from 23% to 46%). One of the main causes of these errors is the spatial variability of fuel consumption within burned areas. This paper studies whether burn severity (BS) maps can be used to improve BE assessment. A burn severity map of two large fires in California was obtained by inverting a simulation model constrained by post-fire observations from Landsat TM imagery. Model output values of BS were validated against field measurements, obtaining a high correlation (R² = 0.85) and low errors (Root Mean Square Error, RMSE = 0.14) throughout a wide range of BS levels. The BS map obtained was then used to adjust BE reference values per vegetation type found in the area before the fire. The adjusted burning efficiency (BEadj) was compared to the burned biomass, which was estimated by subtracting vegetation indices from pre- and post-fire images. Results showed a high correlation for conifers (R² = 0.75) and hardwoods (R² = 0.73), and a moderate correlation (R² ∼ 0.5) for shrubs and grasslands. In general, for all vegetation types BEadj performed better (R² = 0.4-0.75) than literature-based BE (R² < 0.0001). This study demonstrates: (i) the consistency of the simulation model inversion for BS estimation in temperate ecosystems, and (ii) the improvement of BE estimation when the spatial variability of the combustion was quantified in terms of BS.     

Evaluation of a non-point source pollution model,AnnAGNPS, in a tropical watershed

Impaired water quality caused by human activity and the spread of invasive plant and animal species has been identified as a major factor of degradation of coastal ecosystems in the tropics. The main goal of this study was to evaluate the performance of AnnAGNPS (Annualized Non-Point Source Pollution Model), in simulating runoff and soil erosion in a 48 km² watershed located on the Island of Kauai, Hawaii. The model was calibrated and validated using 2 years of observed stream flow and sediment load data. Alternative scenarios of spatial rainfall distribution and canopy interception were evaluated. Monthly runoff volumes predicted by AnnAGNPS compared well with the measured data (R² = 0.90, P < 0.05); however, up to 60% difference between the actual and simulated runoff were observed during the driest months (May and July). Prediction of daily runoff was less accurate (R² = 0.55, P < 0.05). Predicted and observed sediment yield on a daily basis was poorly correlated (R² = 0.5, P < 0.05). For the events of small magnitude, the model generally overestimated sediment yield, while the opposite was true for larger events. Total monthly sediment yield varied within 50% of the observed values, except for May 2004. Among the input parameters the model was most sensitive to the values of ground residue cover and canopy cover. It was found that approximately one third of the watershed area had low sediment yield (0–1 t ha⁻¹ y⁻¹), and presented limited erosion threat. However, 5% of the area had sediment yields in excess of 5 t ha⁻¹ y⁻¹. Overall, the model performed reasonably well, and it can be used as a management tool on tropical watersheds to estimate and compare sediment loads, and identify “hot spots” on the landscape.     

An intercomparison of three remote sensing-based energy balance models using Large Aperture Scintillometer measurements over a wheat-corn production region

This paper compares three remote sensing-based models for estimating evapotranspiration (ET), namely the Surface Energy Balance System (SEBS), the Two-Source Energy Balance (TSEB) model, and the surface temperature-vegetation index Triangle (TVT). The models used as input MODIS/TERRA products and ground measurements collected during the wheat and corn growth period in a subhumid climate at a measurement station in Yucheng, China. MODIS land surface temperature (LST) and leaf area index (LAI) products, corrected using ground-truth observations, were used in the three models. The TSEB model output of sensible (H) and latent (LE) heat fluxes were in good agreement with Large Aperture Scintillometer (LAS)-measured H and LE derived by residual (RMSD < 45 W/m²). Reasonable agreement was also obtained with the SEBS model output yielding RMSD for H of ~ 40 W/m² and LE ~ 55 W/m². However, the TVT model output resulted in poor agreement with the LAS-estimated H and LE with RMSD-values > 110 W/m². Using the uncorrected MODIS LST and LAI products resulted in a deterioration of the agreement in H and LE with LAS-estimated values for both the TSEB and SEBS models, whereas TVT performance improved marginally. These results indicate that the TSEB model yielded the closest agreement with the LAS-estimated fluxes using either the corrected or uncorrected MODIS inputs (LST and LAI). The SEBS model also computed reasonable H and LE values but was significantly more sensitive to errors in MODIS LST and LAI inputs than the TSEB model. In the TVT model, output of H and LE was unacceptable in either scenario of MODIS input which was attributable to errors in selection of the dry edge. With the TVT method, accurate determination of the dry edge end member is critical in regional ET estimation, but for humid and subhumid regions this end member may often be quite difficult to identify or encompass within a satellite scene.     

Estimating northern peatland CO2 exchange from MODIS time series data

Studies using satellite sensor-derived data as input to models for CO₂ exchange show promising results for closed forest stands. There is a need for extending this approach to other land cover types, in order to carry out large-scale monitoring of CO₂ exchange. In this study, three years of eddy covariance data from two peatlands in Sweden were averaged for 16-day composite periods and related to data from the Moderate Resolution Imaging Spectroradiometer (MODIS) and modeled photosynthetic photon flux density (PPFD). Noise in the time series of MODIS 250 m vegetation indices was reduced by using double logistic curve fits. Smoothed normalized difference vegetation index (NDVI) showed saturation during summertime, and the enhanced vegetation index (EVI) generally gave better results in explaining gross primary productivity (GPP). The strong linear relationships found between GPP and the product of EVI and modeled PPFD (R² = 0.85 and 0.76) were only slightly stronger than for the product of EVI and MODIS daytime 1 km land surface temperature (LST) (R² = 0.84 and 0.71). One probable reason for these results is that several controls on GPP were related to both modeled PPFD and daytime LST. Since ecosystem respiration (ER) was largely explained by diurnal LST in exponential relationships (R² = 0.89 and 0.83), net ecosystem exchange (NEE) was directly related to diurnal LST in combination with the product of EVI and modeled PPFD in multiple exponential regressions (R² = 0.81 and 0.73). Even though the R² values were somewhat weaker for NEE, compared to GPP and ER, the RMSE values were much lower than if NEE would have been estimated as the sum of GPP and ER. The overall conclusion of this study is that regression models driven by satellite sensor-derived data and modeled PPFD can be used to estimate CO₂ fluxes in peatlands.     

Remote sensing the orographic effects of dry-hot valley on vegetation distribution in the southeast Tibetan Plateau

ABSTRACT

Mountains in the southeast Tibetan Plateau (TP) often intercept and precipitate abundant monsoon-transported vapours, but some deep valleys of this region are likely subjected to heavy water stress possibly related to orographic effects. Understanding the orographic effects of these dry-hot valleys (DHV) on vegetation distribution is crucial to project local ecological response to global warming. In the study, we used multiple satellite observations with limited in-situ records to investigate the links between vegetation cover and geomorphology in the southeast TP. We designed two types of transects to distinguish altitudinal properties of heat and vegetation between the DHV and non-DHV areas with satellite-retrieved enhanced vegetation index and land surface temperature (LST). Our results showed that the DHVs are characterized by the seemingly ‘abnormal’ decreasing of vegetation density from intermediate elevation simultaneously towards both ridge and valley. The significant increase in LST lapse rate with valley depth (1.8 × 10^?3°C km^?1 m^?1, p < 0.01) suggested the positive role of local valley wind system in the DHV development. Satellite observations revealed that there are, respectively, about 530, 420, and 300 km of DHVs developed in the Nujiang, Lancangjiang, and upper Yangtze rivers, and the DHVs are mostly deeper than 1600 m. Current global warming may lead to the altitudinal expansion of DHV dry and hot effects on local ecosystems, which should be carefully accounted in local ecosystem conservation and management.     

Remote sensing of canopy light use efficiency in temperate and boreal forests of North America using MODIS imagery

Light use efficiency (LUE) is an important variable characterizing plant eco-physiological functions and refers to the efficiency at which absorbed solar radiation is converted into photosynthates. The estimation of LUE at regional to global scales would be a significant advantage for global carbon cycle research. Traditional methods for canopy level LUE determination require meteorological inputs which cannot be easily obtained by remote sensing. Here we propose a new algorithm that incorporates the enhanced vegetation index (EVI) and a modified form of land surface temperature (T_m) for the estimation of monthly forest LUE based on Moderate Resolution Imaging Spectroradiometer (MODIS) imagery. Results demonstrate that a model based on EVI × T_m parameterized from ten forest sites can provide reasonable estimates of monthly LUE for temperate and boreal forest ecosystems in North America with an R² of 0.51 (p < 0.001) for the overall dataset. The regression coefficients (a, b) of the LUE–EVI × T_m correlation for these ten sites have been found to be closely correlated with the average EVI (EVI_ave, R² = 0.68, p = 0.003) and the minimum land surface temperature (LST_min, R² = 0.81, p = 0.009), providing a possible approach for model calibration. The calibrated model shows comparably good estimates of LUE for another ten independent forest ecosystems with an overall root mean square error (RMSE) of 0.055 g C per mol photosynthetically active radiation. These results are especially important for the evergreen species due to their limited variability in canopy greenness. The usefulness of this new LUE algorithm is further validated for the estimation of gross primary production (GPP) at these sites with an RMSE of 37.6 g C m^? 2 month^? 1 for all observations, which reflects a 28% improvement over the standard MODIS GPP products. These analyses should be helpful in the further development of ecosystem remote sensing methods and improving our understanding of the responses of various ecosystems to climate change.     

New refinements and validation of the MODIS Land-Surface Temperature/Emissivity products

This paper discusses the lessons learned from analysis of the Moderate Resolution Imaging Spectroradiometer (MODIS) Land-Surface Temperature/Emissivity (LST) products in the current (V4) and previous versions, and presents eight new refinements for V5 product generation executive code (PGE16) and the test results with real Terra and Aqua MODIS data. The major refinements include considering surface elevation when using the MODIS cloudmask product, removal of temporal averaging in the 1 km daily level-3 LST product, removal of cloud-contaminated LSTs in level-3 LST products, and the refinements for the day/night LST algorithm. These refinements significantly improved the spatial coverage of LSTs, especially in highland regions, and the accuracy and stability of the MODIS LST products. Comparisons between V5 LSTs and in-situ values in 47 clear-sky cases (in the LST range from − 10 °C to 58 °C and atmospheric column water vapor range from 0.4 to 3.5 cm) indicate that the accuracy of the MODIS LST product is better than 1 K in most cases (39 out of 47) and the root of mean squares of differences is less than 0.7 K for all 47 cases or 0.5 K for all but the 8 cases apparently with heavy aerosol loadings. Emissivities retrieved by the day/night algorithm are well compared to the surface emissivity spectra measured by a sun-shadow method in two field campaigns. The time series of V5 MODIS LST product over two sites (Lake Tahoe in California and Namco lake in Tibet) in 2003 are evaluated, showing that the quantity and quality of MODIS LST products depend on clear-sky conditions because of the inherent limitation of the thermal infrared remote sensing.     

Gaseous phase above Ru–O system: A thermodynamic data assessment

The present study is a critical assessment of thermochemical data for gaseous ruthenium oxides based on available experimental data. A full critical analysis and a reinterpretation of data are presented with a proposition for new accurate standard formation enthalpies values: Δ_f H°₂₉₈(RuO₄, g) = −197.6 ± 5.5 kJ mol⁻¹, Δ_f H°₂₉₈(RuO₃, g) = −53.0 ± 10 kJ mol⁻¹, Δ_f H°₂₉₈(RuO₂, g) = 158 ± 20 kJ mol⁻¹ and Δ_f H°₂₉₈(RuO, g) = 301 ± 28 kJ mol⁻¹.     

Embedded long-period fiber grating bending sensor

The use of embedded long-period grating (ELPG) in carbon–fiber composite laminate for bending measurement has been demonstrated in this paper. As the bending curvature increases on the ELPG laminate, the coupling strength of the cladding mode decreases while the resonance wavelength remains relatively constant. The ELPG yields a sensitivity of 5.065 dB m⁻¹ and repeatability of 98.1% up to a bending curvature of 2 m⁻¹. It also can be used to determine direction of the bend.     

Development of an optical fibre reflectance sensor for copper (II) detection based on immobilised salicylic acid

Immobilized salicylic acid onto XAD-2 (styrene–divinylbenzene cross-linked copolymer) has been attempted in this study as a reagent phase for the development of an optical fibre copper (II) sensor. The measurements were carried out at a given wavelength of 690.27 nm since it yielded the largest divergence different in reflectance spectra before and after reaction with the analyte element. The optimum response was obtained at pH 5.0. The linear dynamic range of Cu(II) was found within the concentration range of 1.0–2.0 mmol L⁻¹ with its LOD of 0.5 mmol L⁻¹. The sensor response from different probes (n = 9) gave an R.S.D. of 8.4% at 0.55 mmol L⁻¹ Cu(II). The effect of interfered ions at 1:1 molar ratio of Cu(II):foreign ion was also studied in this work.     

Thermodynamic assessment of RuO4 oxide

RuO₄ oxide appears much less stable than RuO₂(s) in the Ru–O binary system with a melting point close to room temperature and a certain propensity to vaporize or decompose at low temperatures. Ab initio simulations in the framework of density functional theory (DFT) on RuO₄(s) are performed to analyze the cubic and monoclinic structures and to evaluate the heat capacities at low temperatures. Then, a critical evaluation of thermodynamic data from calorimetry and vapor pressure determinations - was carried out coupled with ab-initio calculations to propose new thermodynamic data: the entropy.S° (RuO_4, s, cubic, 298K) = 132.7 J·K⁻¹mol⁻¹ and formation enthalpy.Δ_fH° (RuO₄, s, cubic, 298K) = −252.4 ± 5.5 kJ mol⁻¹.     

A 24-h forecast of ozone peaks and exceedance levels using neural classifiers and weather predictions

A neural network combined to a neural classifier is used in a real time forecasting of hourly maximum ozone in the centre of France, in an urban atmosphere. This neural model is based on the MultiLayer Perceptron (MLP) structure. The inputs of the statistical network are model output statistics of the weather predictions from the French National Weather Service. These predicted meteorological parameters are very easily available through an air quality network. The lead time used in this forecasting is (t + 24) h. Efforts are related to a regularisation method which is based on a Bayesian Information Criterion-like and to the determination of a confidence interval of forecasting. We offer a statistical validation between various statistical models and a deterministic chemistry-transport model. In this experiment, with the final neural network, the ozone peaks are fairly well predicted (in terms of global fit), with an Agreement Index = 92%, the Mean Absolute Error = the Root Mean Square Error = 15 μg m⁻³ and the Mean Bias Error = 5 μg m⁻³, where the European threshold of the hourly ozone is 180 μg m⁻³.To improve the performance of this exceedance forecasting, instead of the previous model, we use a neural classifier with a sigmoid function in the output layer. The output of the network ranges from [0,1] and can be interpreted as the probability of exceedance of the threshold. This model is compared to a classical logistic regression. With this neural classifier, the Success Index of forecasting is 78% whereas it is from 65% to 72% with the classical MLPs. During the validation phase, in the Summer of 2003, six ozone peaks above the threshold were detected. They actually were seven.Finally, the model called NEUROZONE is now used in real time. New data will be introduced in the training data each year, at the end of September. The network will be re-trained and new regression parameters estimated. So, one of the main difficulties in the training phase – namely the low frequency of ozone peaks above the threshold in this region – will be solved.