Three prizes from the European Meteorological Society were awarded to Valentina Sicardi (MPI of Biogeochemistry, Germany) Ana Prieto-Blanco (University of Wales, UK), and Raul Lopez-Lozano (CITA, Spain) for the three best posters.
The LANS-Alpha Model of Sub-Grid Scale Turbulence in the POP Ocean Model
Los Alamos National Lab.
POP, the Parallel Ocean Program developed and maintained at Los Alamos National Lab., is widely used by the ocean and climate modeling community. Like all numerical models, computational time limits the spatial resolution at which POP can operate; standard simulations use grids of 0.5 to 1 degree in latitude and longitude. This resolution does not capture the motion of eddies at the Rossby radius of deformation, and thus lacks the correct energy cascade and heat transport at these scales. The Lagrangian-Averaged Navier-Stokes alpha (LANLS-alpha) model, developed by Darryl Holm and colleagues at LANL, improves these characteristics with a smoothed advecting velocity and an additional nonlinear term.
Preliminary results show that the POP-alpha model improves measures which depend on the resolution of meso-scale eddies, such as vertical temperature profile, kinetic energy, and eddy kinetic energy. In some cases these improvements are comparable to a doubling of horizontal resolution, which increases computation by a factor of ten, while the addition of the alpha model parameterization only adds a few percent in computational time.
Influence of El Nino on the Biennial Rossby Waves in the Indian Ocean with a Special Emphasis on Indian Ocean Dipole
Bakshi Hardeep Vaid
The interannual variability of the tropical Indian Ocean is examined using 44 years (1958 – 2001) of Simple Ocean Data Assimilation (SODA) sea surface height anomalies (SSHA) and Hadley centre Ice sea surface temperature anomalies. SODA SSHA over the Indian Ocean is filtered using two-dimensional Finite Impulse Response filter and the filtered biennial Rossby wave component is analyzed to understand its variability in the tropical Indian Ocean. From the composite map of the biennial Rossby wave signals, a dipole pattern is clearly observed in the equatorial Indian Ocean with the wave crest occupying the western Indian Ocean and the wave trough occupying the southeastern tropical Indian Ocean during the positive Indian Ocean Dipole (IOD) events. Downwelling biennial Rossby waves along 1.5oS are seen propagating westward from the eastern boundary, more than one year prior to the formation of a positive IOD and reach in the western equatorial Indian Ocean during the peak positive dipole time. Influence of El Nino on the propagation of biennial in the Indian Ocean has been observed. In short, the present study supports the intrinsic link between IOD and El Nino.
Advances in Coastal Altimetry over the Northwestern Mediterranean
LEGOS - CTOH/POC
We describe the outcomes of research into applications of altimetry in the Northwestern Mediterranean (hereafter NWM) Sea for the following medium term purposes: to contribute to the oceanographic knowledge of the area and to assess the applicability of altimeter techniques in coastal systems.
The region is a perfect site for such investigations in virtue of availability of various long-term in situ observations.
Altimetry is well suited to the study of the general circulation in the open ocean. However using altimetry is still a challenging issue for coastal studies such as in the NWM partly because the dynamic is much more complex. Thus, an important aspect of the research activity consisted in assessing the extent to which altimetry can be employed in the area, and devising any possible improvement in the data processing chain and in the retrieval of the geophysical parameters in order to make altimetry more applicable. We describe the adopted processing strategy and the comparison of the retrieved parameters with in situ data. The scales accessible through the multi-sensor observing systems are then compared with the 3D coastal model Symphonie in order to show their capabilities to improve our knowledge of the coastal area in the NWM Sea and especially concerning the Catalan-Ligurian-Provençal current variability.
Effects of Absorbing Aerosols in the SST Using AI (TOMS) and AOT (SeaWiFS)
Ana Belén Ruescas Orient
The presence of aerosols affects the measurement of sea surface temperatures (SST), extracted by means of sensors situated on satellites, because they are capable of absorbing and dispersing infrared radiation. Dust particles, smoke, and volcanic ash make up what is called atmospheric aerosols, and are key elements in cloud formation and in precipitation. It is even believed that aerosols with high iron content are crucial for the growth of phytoplankton in the sea (Bigg, 2003). In this study, reference will be made to only one type of "naturally” occurring aerosol, that is, one not originating from anthropogenic emissions, like dust from deserts. It will specifically deal with the influx of dust from the Sahara desert and the deserts of the Arabian Peninsula into the environment around the Mediterranean Sea.
This influx leads to serious errors in the extraction of SST through satellite imagery, which makes it necessary to carry out both a detailed study of the effect they have on absorption for the channels in infrared wavelengths, as well as a proposal of methods for correcting these errors.
The main objective of this study is to evaluate the errors caused by these aerosol pulses in the SST extracted by the AVHRR sensor through the use of two indicators of aerosol presence: the TOMS Earth Probe aerosol index (AI) and the Aerosol Optical Thickness (AOT) derived from SeaWiFS.
Earth Observation data assimilation in marine forecasting
Lars Boye Hansen
Marine forecasting of physical and biological variables is increasingly being used to assist decision making by authorities and industry. For this purpose, data assimilation is an ideal framework for constraining the models by measured marine variables. Earth observation sea surface temperature (SST) and chlorophyll-a data provide a unique data set in this respect. It has a high spatial resolution, but introduces the challenge of handling information that is patchy in space and irregular in time.
In this contribution examples of operational forecast models for the Sea of Chiloé and the inner Danish waters are presented. Near real time SST and and chlorophyll-a data are fed into the system via a simplified Kalman Filter approach, in which the dynamical model is basically performing an intelligent patch interpolation of the measured variables. Prior to the assimilation the EO data are interpreted by use of the model variables themselves and an additional surface layer turbulence description, thus transforming the measurements into the model space. The performance of the EO assimilation approach is compared to the existing operational services at both locations.
Modelling the shelf recruitment of c. finmarchicus on the west coats of Norway
Mohn-Sverdrup Center / NERSC
The zooplankton species c. finmarchicus is important in the Norwegian Sea because it is the main food for planktivorious fish such as herring and capelin and for larvae of most commercial fish stocks. c. finmarchicus has a one-year life-cycle, and the shelf population is recruited each spring from individuals that has over-wintered at depth either in a fjords or in the deep basins in the Norwegian Sea. The shelf recruitment of c. finmarchicus is investigated using a numerical ocean model, the HYbrid Coordinate Ocean Model (HYCOM), that forces an individual based model (IBM) for c. finmarchicus. The model is set up on a 4 km grid along the west coast of Norway and the focus is on shelf recruitment from the Norwegian Sea population. Numerical experiment are performed investigate the effect of wind, temperature, food availability on the size and fitness of the shelf population. Here, the focus is on the years 1995 and 1996; 1995 is characterized by many storms and strong southwesterly winds, whereas 1996 is characterized bye fewer storms and weaker southwesterly winds. The shelf recruitment under these two very different wind-regimes is investigated.
Issues about retrieving sea surface salinity in coastal areas from SMOS data
Laboratoire d'Océanographie et du Climat - Expérimentation et Approches Numériques (LOCEAN)/Université Pierre et Marie Curie-Paris 6/CNRS
Ocean salinity is a key parameter in oceanic and climate studies. Together with the ocean temperature, the salinity influences the density of the water masses and actively participates in their formation and circulation. In situ sea surface salinity (SSS) measurements, done by buoys and oceanographic or commercial ships, remain sparse and irregular, with large parts of the global ocean never sampled. In order to fill this gap, two missions carrying L-band (1.4 GHz) radiometers have been recently proposed: the Soil Moisture and Ocean Salinity (SMOS, ESA) and Aquarius (NASA) missions. Their objective is to estimate the SSS on a global scale, in 200 km x 200 km boxes on a 10-day average with a precision of 0.2 psu (practical salinity units, corresponding to parts per thousand). They should allow a global monitoring of the SSS at a synoptic scale, and the remotely sensed SSS should be suitable for assimilation into ocean circulation models, according to GODAE requirements.
This study deals with SSS retrieval from SMOS measurements in coastal zones. These zones are characterized by strong and variable SSS gradients (several psu) on relatively small scales. In particular, the extent of river plumes is highly variable, typically at kilometric and daily scales. Monitoring this variability from L-band satellite radiometric measurements is particularly challenging because of the resolution of the satellite measurements (typically 30-100 km) and because SMOS measurements over the coastal ocean are contaminated by the nearby land: land brightness temperatures range from approximately 200 to 300 K, compared to approximately 100 K for the ocean.
The objective of this study is to assess to which extent SSS variability in coastal areas can be monitored with SMOS. On the one hand, a set of academic tests are conducted with a linear coastline and constant geophysical parameters. On the other hand, more realistic tests are conducted over the Bay of Biscay, using SSS and SST fields simulated by the MARS3D model developed by the IFREMER. SMOS brightness temperatures are deduced from the measured visibilities using image reconstruction. The impact of different apodization windows on the retrieved SSS as a function of the distance to the coast is investigated. Under SMOS configuration, an ocean grid point is seen by several cells with various view angles, various resolutions and various ellipticities. The effects of the cells maximum resolution and elongation are investigated by studying the bias and error on the retrieved SSS with respect to the distance to the coast.
Observing Ocean Mass and Heat Storage Changes using GRACE and Altimetry Data
TU Delft - DEOS
The Gravity Recovery And Climate Experiment (GRACE) monthly solutions of the time-variable gravity field allows direct estimates of changes in the ocean water mass budget. In combination with satellite altimetry observations, this can be used to estimate heat content changes in the ocean as well. Since the amplitudes of these signals are relatively small (compared to the signal over land), it is important that both the GRACE and altimetry data are corrected and combined in a consistent manner.
In this presentation we focus on two corrections, i.e. the geocenter correction and de-aliasing of the altimetry data for high-frequency signals. The former is due to the fact that the GRACE satellites are insensitive to variations of the center of mass of the Earth, whereas altimetry is not; the latter applies to oceanic signals with periods shorter than the Nyquist period of the altrimetry observations. Additionally, we discuss other corrections that should be applied to correctly reconcile the two data sets and present the results.
SSALTO CALVAL Performance assessment Jason-1 GDR "B" / GDR "A"
Altimeter Jason-1 data (GDR: geophysical data record) were processed in version A until cycle 135 and in version B from cycle 136 until current cycle.
Recently, a reprocessing of the GDR was done by JPL (cycles 1 to 21) and CNES (cycles 128 to 135). The main evolutions in the GDR 'B' are the implementation of a new retracking algorithm (order 2 MLE4), a new precise orbit based on a GRACE gravity model and new geophysical corrections (tidal models, MOG2D, Sea State Bias).
The objective of this study is to compare the 'A' and 'B' versions of the GDRs. The impact of each change in GDR 'B' is analyzed as well as the global impact on Sea Surface Height (SSH) performances.
Satellite chlorophyll as a tracer for upward velocities in the surface ocean
Upward velocities have been diagnosed through the development of an approach based on the combination of remotely sensed daily images of both chlorophyll and temperature at the ocean surface in subtropical latitudes (Gulf of Cádiz and Alboran Sea). This novel approach allows to detect areas of vigorous vertical velocities, through the generation of a calculation algorithm built from the simple Lagrangian model of phytoplankton growth that occurs during the rise of deep waters to surface.
Cold waters poor in chlorophyll are quite unusual during spring-summer at the surface of the subtropical latitudes. They result from very intense dynamic processes that lift deep water of different properties to surface. Detection of these waters by radiometers in orbit can be used as a satellite tracer for the areas where these processes occur. The fast initial evolution of chlorophyll and temperature in surface limits the implementation of the algorithm to high upward velocities. Assuming this limitation, the values of vertical velocity diagnosed have been found to be coherent with independent estimations based on diffusive and advective inputs of nitrogen into the euphotic zone in the area.
The Mercator eddy permitting global ocean forecasting system
The Mercator-Ocean eddy permitting global ocean forecasting system has a 1/4° horizontal resolution and 46 levels on the vertical (ORCA025), and is forced with daily surface fluxes from ECMWF operational analyses. Satellite Sea Level Anomaly measurements (JASON, ENVISAT and GFO) from january 2005 up to real-time are assimilated with a reduced order optimal interpolation scheme (ROOI). Due to its relatively fine resolution, the forecasting system provides an integrated description of the ocean with a realistic description of the meso scale features.
Assimilation scores are presented, and independant in situ data of the Atlantic, Pacific and Antarctic ocean basins are compared with the simulation results in order to provide an estimation of the performances of the system. The results are also compared with the Levitus climatology, and with the ARMOR weekly products, which optimally combine satellite (SST, SLA) and in-situ (T/S profiles) near real time observations.
First results of the next generation of Mercator ocean forecasting systems are presented, which use a multivariate and multidata assimilation method, and which comprise a sea ice model (LIM). This system will replace in 2007 the current operational system, and will allow us to perform a global ocean reanalysis for the 1992-2005 period.
CryoSat Validation Experiment CryoVEx 2006 at Lincoln Sea
Finnish Intitute of Marine Research
CryoSat Validation Experiment CryoVEx 2006 was carried out in April and May 2006. Part of the experiment was made on the sea ice of Lincoln Sea. This experiment was a joint venture between European Space Agency (ESA), Danish National Space Center (DNSC), Alfred Wegener Institut (AWI), Scottish Association for Marine Science (SAMS) and Finnish Institute of Marine Research (FIMR).
Due to very unfortunate launch failure of CryoSat satellite, aircraft-borne radio altimeter ASIRAS was used in the experiment. ASIRAS is similar to SIRAL instrument to be flown onboard CryoSat II. Ice thickness measurements were also made with laser altimeter and a helicopter-borne electromagnetic ice thickness sensor known as EM-bird.
During the airborne measurements ground teams collected in situ data including ice thickness, density, salinity and snow properties. Overall description of Lincoln Sea experiment including planning and realization is presented. Preliminary comparisons between in situ and remote sensing data are also presented.
The Rhine Region of Fresh Water Influence
Gerben De Boer
Delft University of Technology
Co-authors: J.D. Pietrzak, J.C. Winterwerp
In the Rhine ROFI (Region Of Freshwater Influence) stratification significantly affects tidal currents. Cross shore velocities and stratification signals are known to exhibit dominant semi-diurnal and fortnightly variations due to tidal advection and mixing. We investigate the semi-diurnal variation known as tidal straining. First, we used the 15-year 1-km NOAA-SST dataset at KNMI. We selected time series that have over three images/day, do not coincide with strong wind and show a marked temperature contrast between stratified water and the surrounding sea caused by solar heating. The timing of the observed upwelling, down welling and plume displacement vector are in accordance with tidal straining theory. Second, an idealized 3D numerical model of the Rhine ROFI was employed to explore the effect of stratification on the vertical structure of tidal currents. In line with observations it shows anti-cyclonically rotating surface currents and cyclonically rotating bottom currents. As a first approximation, this can be understood with the dynamic Ekman equations (Prandle, 1982). In the special Kelvin wave case (zero depth averaged cross shore currents) it already predicts a transition from almost rectilinear tidal currents to an ellipticity veering of 50% as a function of bulk vertical eddy viscosity only.
Envisat Ocean applications
Cristina Martin Puig
Starlab Barcelona S.L.
The uncontrolled energy demand and the strong dependence on fossil fuels, have encourage various energy sectors to work together in order to reduce the pollution emission, while preserving the equilibrium of the ecosystems and the sustainable economic development. Renewal energies are an efficient solution to this concern, as well as the motivation of EOLISCAT; an internal Starlab Barcelona S.L. project which aims to retrieve mesoscale wind fields from ASAR data. The results acquired will help to the assessment of offshore wind farms deployment along the Catalan coast, and widely contribute to the use of renewal energies in Spain. The poster will describe the research work performed at Starlab, as well as preliminary results.
ASAR is not the unique Envisat sensor that may help industry development; multispectral satellite sensors such as Envisat's Medium Resolution Imaging Spectrometer (MERIS) can observe the colour of surface water as a means of deriving their environmental quality. Within this framework, the use of MERIS data to support water bodies’ sustainability reports will be also presented in this poster.
Introduction of inequality constraints into a Reduced Order Kalman Filter for data assimilation into ocean circulation models
In data assimilation into oceanographic models, one of well-known classical problems is the estimation of variables physically unrealistic. The aim in this work is to take into account inequality constraints within a reduced order Kalman filter (the SEEK filter) to improve actual method assimilating remote sensing data of temperature, salinity, pressure and altimetry. These inequality constraints are specific to the hybrid ocean circulation model used (HYCOM) which needs all layer thicknesses positive as well as the density increasing with depth. To follow these constraints within the assimilation scheme, a quadratic programming problem is inserted after each analysis step. The model with an example of violated constraints after a SEEK analysis are presented. Then, the purposed approach SEEK-3DVar is explained theorically with discussion on several resolution methods. First results are finally presented.
Estimates of Geostrophic Transport in the South Atlantic Bight Using Satellite Altimetry
University of South Carolina
Accurately quantifying long-term Gulf Stream transport is critical to understanding the meridional heat transfer cycle and associated global climate. In this study, in situ current measurements from cross-stream transects of an Acoustic Doppler Current Profiler were used in conjunction with satellite altimetry data to construct a model to predict Gulf Stream volume transport. Additionally, CTD profiles were used to estimate the geostrophic velocity via the thermal wind equation. The measured velocity profiles were compared to the calculated velocity profiles using the thermal wind equation to produce a statistical model of vertical variability in Gulf Stream flow. Depth-independent estimates of the geostrophic current were computed from cross-stream sea surface height anomaly data (SSHA) obtained from the TOPEX/POSEIDON and Jason-1 satellite altimeters. These estimates provide the boundary condition for the flow estimates obtained from the thermal wind equation calculations. A data-calibrated model emerged that generated a decadal time series of Gulf Stream transport for this area (31.5°N 79°W) of the South Atlantic Bight (SAB) and could accurately compute the local depth-integrated volume transport using historical SSHA data. Gulf Stream transport was observed to increase gradually, not dramatically, in the SAB from the Straits of Florida to just upstream of the Charleston Bump.
Applications of Satellite Oceanography
Satellite data are widely used within the field of commercial oceanography in the assessment of winds, waves and currents. The first example illustrates the use of data sources such as Topex/Poisedon, JASON and ERS-1 and 2 in the calibration of a global wind and wave model and its subsequent use in projects such as wave energy resource mapping and near-shore wave modelling. The second example discussed is the number of recent of studies into the tracking of large scale surface features such as the Gulf of Mexico Loop Current, the North Brazil Current and the Aghulus Current. Combining altimetry data and hindcast numerical model data have enabled these systems to be effectively modelled and forecast enabling an assessment of their past and potential impact on drilling locations.
Sea oil slicks remote sensing using MERIS spectral bands
In this work are presented the results obtained by simulating the total remote sensing reflectance at nadir for oil in sea water, correlated with MERIS bands. The difference between water radiance (taking into account also chlorophyll) and oil radiance has been implemented into a contrast model that allows to verify the possibility to detect oil slicks in sea water by means of MERIS images.
Water quality monitoring of the Eastern part of Gulf of Finland.
Saint-Petersburg Center for Hydrometeorology and Environmental Monitoring
Due to many natural and antropogenic factors Eastern part of Gulf of Finland is extremely sensitive to human impact. In order to manage local water resources well with minimum risks for the nature having precise online data about environmental state is vitally important.
The aim of the research is to realize water quality monitoring in the area of Gulf of Finland based on use of the data from the Moderate Resolution Imaging Spectroradiometer (MODIS/EOS Terra) and radiometer AVHRR/NOAA. One of the main goals of the research is to conduct an investigation wether the aforesaid type of sattelite data is usable for researching such quantative and qualitative charachteristics of water in the area of Gulf of Finland as sea surface temperature, concentration of chlorophyll »a« and water turbidity which allow to trace thermal anomalies of the surface waters, algal blooms and concentration of the suspended solids. Results of satellite data processing are verified according to the ship observations and data from the hydrometeorological ground-based observation stations network.
The Urban energy balance: Parameterization of the anthropogenic heat flux
Katholieke Universiteit Leuven
Although many studies show that heat released from human activities (anthropogenic heat Qa) can be a significant contributor to the urban energy balance, QA is often not well parameterized or even disregarded in atmospheric modeling. Often, the ground surface flux and anthropogenic heat fluxes are combined and calculated as residual of the energy balance, thereby incorporating all errors, both in measurements and the model. Therefore, the goal of this study is to explicitly parameterize the anthropogenic heat released from industry, traffic and domestic households based on a spatial distribution of annual energy consumption (in France), by means of land use and population density maps. A time dissagregation is obtained using the concepts of ‘heating degree-days’ and ‘time factors’ for building heating and traffic respectively. A 120x120 km² domain centered above Marseille is selected to perform runs with the mesoscale meteorological model ARPS during a 3-day IOP (20 – 22 June 2001). Effects of QA on the urban energy balance are investigated using eddy covariance observations in the center of Marseille provided by the ESCOMPTE campaign (Expérience sur Sites pour COntaindre Modèles de Pollution et Transport d’Emissions). Contrasts with rural sites are found using 2 rural stations Meyrargues (maize) and Trets (grass).
Measuring Carbon Dioxide from Space
University of Leicester
Until recently, Earth based measurements of CO2 concentrations have been sparsely distributed, and have poorly represented continental land mass, where localised sources and sinks overwhelm measurements. SCIAMACHY and later the Orbiting Carbon Observatory data will be used to give a more accurate quantification and understanding of the carbon sources and sinks. It is a requirement of the Kyoto protocol to accurately quantify and monitor CO2 sources and sinks.
The advanced FSI-WFM –DOAS method has been developed to retrieve total columns of greenhouse gases (primarily CO2) with increased precision, enabling the investigation of:
• The accuracy of global CO2 measurements from NIR observations of SCIMACHY for the direct determination of flux estimates.
• Co-retrieval methods as a way of assessing emission fluxes.
• The application of cloud resonance as a method of assessing the sensitivity to PBL CO2.
• Data assimilation of CO2 measurements as a means of providing a better error estimate for CO2 retrieval errors and fluxes.
• Quantification of greenhouse and related biospheric gas fluxes through inverse modelling constrained by assimilation of ENVISAT measurements, with a focus on CO2.
Inverse modeling of emissions at a local scale: a study of the Paris area
McGill University / Meteorological Service of Canada
For chemistry-transport models operated in atmospheric photo-oxidant pollution at a local scale, surface emissions are an essential input data to which output concentrations are very sensitive. Emissions are nevertheless poorly known and remain a major source of uncertainty in current models. The possibility of modifying existing inventories seems therefore promising for a better simulation of concentration fields and a better understanding of pollution.
Emissions of the Paris area are optimized with a methodology for the inversion of surface anthropogenic emissions at a local scale based on the chemistry-transport model CHIMERE and its adjoint. It uses a kriging technique to optimize the use of the information available in network measurements. A dynamical spatial aggregation technique is elaborated for the Paris area to reduce the size of the problem.
NOx emissions from the inventory elaborated by AIRPARIF were inverted during the summers of 1998 and 1999: events of the ESQUIF campaign are studied in detail. The optimization reduces large differences between simulated and measured concentrations. Generally, the confidence level of the results decreases with the density of the measurement network. Therefore, the results with the higher confidence level correspond to the most intense emission fluxes of the Paris area.
Satellite Characterization of Power Plant Aerosol Emissions
University of Évora
The aim of this work is the identification and characterization of aerosols plumes emitted from the Portuguese Electrical Company Power Plants using satellite measurements, with the purpose of monitoring the emissions of pollutants and studying their atmospheric dispersion. The methodology is based on radiative transfer calculations in the atmosphere that, properly combined with satellite measurements, allow for the detection and determination of the aerosols amount in the atmosphere and some of their physical characteristics. The columnar amount of pollutants obtained in this way (Aerosol Optical Thickness-AOT) is compared with the particles emissions that are monitored on the top of the power plant towers, in order to validate the values retrieved from the satellite-based method.
Satellite measurements from the MODerate Resolution Imaging Spectroradiometer (MODIS) were used, since they present the adequate spectral channels and spatial resolution to observe and monitor disturbances in the Earth-atmosphere system.
The analysis of the AOT values obtained with the satellite-based methodology, confirmed that a higher value of AOT exists in the Power Plants areas and a good agreement can be observed between the values of the optical thickness obtained from the satellite data and the aerosol emissions measured at the top of the power plant towers.
Impacts of Urbanization on the CaRBon Sequstration in Europe
Max-Planck Institute for Biogeochemistry
Numerical simulations with the mesoscale weather-predicting model MM5 suggested that conversion of natural land cover to urban one caused reduction in temperature diurnal range and in precipitation over the areas of perturbation. The simulations showed an average reduction of the diurnal temperature range in regions with urban land cover modification by 1.1°C and 0.9°C in summer and in wintertime respectively. The total amount of precipitation averaged over the whole domain over land was by -2.1 mm month-1 and -2.5 mm month-1 in summer and wintertime respectively. Urban sites exhibited decreased rainfall in summer (-4 mm month-1) and increased rainfall in winter (+3 mm month-1).
This study addresses the question: how do the changes in diurnal temperature range and precipitation associated with urbanization influence the vegetation growth in Europe? Series of numerical simulations with the Biome-BGC terrestrial ecosystem model were performed to determine how productivity of vegetation responded to urban-caused changes in climate of Europe. It was found that ecosystems situated in dry warm environments were the least sensitive to the urban anomaly temperatures and precipitation while the ecosystems in colder and moist climates showed the greater response.
Aerosol Forecasts for Solar Energy Applications
In order to raise the effectiveness of renewable energy production and to integrate the growing solar energy sector into existing power supply systems, reliable near-real-time predictions of solar irradiance are needed for 1-3-day forecasts of energy yields and consumer demands.
Accurate information about clouds, aerosols and water vapour is necessary to calculate ground level irradiance. While water vapour forecasts are already performed with classical weather models, prediction of aerosol distribution is still a matter of research. Especially in mostly cloud-free regions, which are of special interest for larger solar energy facilities, aerosol forecasts are of great importance to obtain accurate irradiance predictions. A first approach is given by the MM5-based EURAD model (European Air Pollution Dispersion Model), developed by the University of Cologne for air quality monitoring purposes. It predicts aerosol distributions distinguishing between components such as soot, organic particles and inorganic acids.
In a case study forecasted AOD of the EURAD system was validated against ground based aerosol measurements for five months in 2003. Results show a slight underestimation of AOD, especially during the summer months. However, deviations remain within currently accepted errors in satellite retrievals of aerosol properties and within accuracy requirements for input parameters of irradiance calculations.
Two-dimensional modelling of MIPAS trace species information content
Belgian Institute for Space Aeronomy
Operational processors of Envisat Michelson Interferometer for Passive Atmospheric Sounding (MIPAS) retrieve temperature and trace species profiles from limb radiance measurements assuming one-dimensional atmosphere. Averaging kernels associated with the retrieval enable one-dimensional characterisation of the retrieved information as a function of altitude or pressure. Nevertheless, the MIPAS limb scanning strategy, concurrent emission and absorption processes, and the orbital progression of Envisat, combine to spread the actually probed information in both the vertical and horizontal (angular) dimensions. Here, we describe a simple radiative transfer model capable of calculating MIPAS limb radiance emission spectra in a two-dimensional atmosphere. The model is applied to MIPAS particulars (selected spectral microwindows, global fit retrieval approach) in order to characterise the vertical and angular distribution of the information available after a complete limb-scanning sequence.
Global Atmospheric CO2 Remote Sensing Data Assimilation
Max Planck Institute for Biogeochemistry
The main objective of this research, on global atmospheric CO2 remote sensing data assimilation, is to develop and apply techniques for the assimilation of remote-sensing and in situ based observations into a global atmosphere/ocean carbon cycle model framework.
A case study with 1DVAR
Tien Du Duc
Hanoi University of Science
HRM (a High resolution Regional Model, developed by DWD) has been running operationally in Vietnam since 2002. Observation should be assimilated to improve the quality of initial fields for this model. Satellite data is the main source for the poor-observation area, especially in sea region. In this study, 1D-vartional assimilation system (developed in Met. Office of U.K) will be applied for HRM with NOAA 15,16 data.
Estimating Regional CO2 Surface Fluxes In Complex Landscapes
Stephan De Wekker
Mountain forests represent a large portion of gross primary productivity within the United States and a significant potential net CO2 sink. Therefore, there is a need to develop methods to estimate regional fluxes of CO2 in mountainous terrain. We present results from a combined modeling and observational study of regional CO2 fluxes in mountainous terrain and discuss the major challenges. We use data from the Airborne Carbon in the Mountains Experiment (ACME), conducted in May and July of 2004. Applying a simple budget method to the aircraft data, we estimated CO2 drawdowns of several ppm in the mountain boundary layer, representing significant CO2 uptake by the forests. To understand the observations, we use a modeling framework consisting of the Regional Atmospheric Modeling System (RAMS) and its adjoint. We prescribe various scenarios of a CO2 flux at the surface and atmospheric conditions resulting in a variety of spatial and temporal behaviors of CO2 concentration in and above the mountain boundary layer. This enables the calculation of surface CO2 fluxes using the same approach as in the observations. Ideas to use the adjoint in a variational data assimilation approach to estimate regional CO2 fluxes will also be presented.
Spatial and Temporal Variation of Rain Fields in Tuscany
University of Pisa
The poster describe our current work in progress at INFN-Pisa, regarding the spatial and temporal inhomogeneity of rain fields using meso scale model (MM5) and fractal geometry.
The objective is to provide valuable information on structures inside the rain fields, such as size (diameter), shape and orientation of rain cells. This information is needed to improve our knowledge on thunderstorms dynamics and to accurately model and simulate the spatial variation of 2-D rain fields.
Spectral coupling method for limited area models
National Meteorological Administration
A one-way nesting method for joining a large scale model with a high resolution limited area model (LAM) has been developed and tested in terms of the spectral information transmission through lateral boundaries. The spectral coupling method based on the bi-Fourier representation of the fields was used as additional step to the presently used flow-relaxation method (Davies, 1983). We investigated the proposed method's capability to supply to the LAM's solution the possibly missing large scale information. The performance of the LAM using spectral coupling versus operational grid-points coupling was analyzed, the results showing a reduction of errors in the forecast and an improved ability to simulate extreme events produced through scales interaction.
Intercomparison of the primal and dual formulations of variational data assimilation
Amal El Akkraoui
The variational formulation of the data assimilation problem can be cast either in its primal or dual formulation.
The primal formulation uses a control variable in model space: it corresponds to the so-called 3D-Var algorithm, whereas the dual formulation proposes a different angle by using a control variable defined in observation space: it corresponds to the Physical-space Statistical Analysis System (PSAS). In the case where the 3D-Var cost function is purely quadratic, both methods have been shown to be theoretically equivalent.
This work compares dual and primal formulations in an operational framework. A PSAS scheme has been formulated using the operational 3D-Var of the Meteorological Service of Canada. Our results show that the two formulations give exactly the same results at convergence, and confirm the analysis presented in Courtier (1997) : that is with their own conditioning, 3D-Var and PSAS require a comparable number of iterations and a similar overall cost to converge.
Finally, the converging properties of both algorithms are looked at when approximate forms of the Hessian matrix by its leading eigenvectors are used to precondition the problem. In PSAS, the Hessian is relatively sparse which makes the preconditioning more efficient. This is attractive particularly in view of the extension to 4D-Var and 4D-PSAS for which a reduction in the number of iterations can make a huge difference in the time required to produce an operational analysis.
Fast cloud properties retrieval scheme with SEVIRI images
Royal Meteorological Institute
In order to deliver near real-time estimates of the top of the atmosphere (TOA) radiative fluxes from the Geostationary Earth Radiation Budget (GERB) broadband radiometer on board of the Meteosat-8 satellite, a radiance-to-flux conversion needs to be performed on the measured radiances. As such a conversion is done by using the angular dependency models (ADMs) developed from the Clouds and the Earth's Radiant Energy System (CERES) experiment, the GERB ground segment has to rely on a scene identification based on SEVIRI data.
In this poster, we present the method developed for the GERB/SEVIRI scene identification within the RMIB GERB Processing (RGP) to estimate the cloud properties needed to properly select the ADM according to the observed scene. This cloud properties retrieval scheme based on reference composite TOA clearsky reflectances allows to estimate robust a cloud flag, cloud optical depth and cloud thermodynamic phase for each SEVIRI pixel. Even if the method is based on lookup tables of radiative transfer computations, its simple non-iterative behaviour allows to perform the retrievals faster than more complex schemes found in the literature with a good confidence.
Impact of a Flow-Dependent Background Error Covariance Model Based on Sensitivity Functions in a 3D-VAR
UNIVERSITÉ DU QUÉBEC À MONTRÉAL (UQÀM)
The a posteriori sensitivity functions estimated as in Laroche et al. (2002) characterise the change in the initial conditions that will lead to an important modification in the short-term forecast. To take into account the specificities of the flow characteristics, those sensitivity functions are introduced as structure functions within the background error covariance matrix of a 3D-Var assimilation system to allow the analysis to fit the observations while at the same time imprint a structure that can trigger the development of a weather system. This approach is referred to as the adapted 3D-Var was first proposed by Hello and Bouttier (2001). A different formulation has been proposed and implemented within the 3D-Var operational system of the Meteorological Service of Canada. Experiments were carried out with different definitions of the sensitivity functions to show their impact on the forecast. The analyses obtained with the adapted 3D-Var were compared with respect to those of the operational 3D-Var and to the sensitivity analysis in terms of their impact on forecasts. The adapted 3D-Var is shown to reduce the forecast error over the targeted area while at the same time improving the fit to the observations over the sensitive areas.
A new kind of Static Fourier Transform Spectrometer for atmospheric CO2 monitoring
CO2 is the major contributor to global warming. Yet, the current knowledge of the carbon cycle cannot permit to predict correctly CO2 influence on the climate. To increase this knowledge the spatial resolution of CO2 concentration measurements must be better. The scientific community has expressed the need of measurements at regional scale (the current on ground measurement network is a continental network). Such a measurement cannot but be achieved thanks to a space remote sensing monitoring. However, measurements must reach a precision better than 1 % on the column mixing ratio in order to provide complementary information to the on ground sparse but accurate measurement network.
CNES has developed a new kind of Static Fourier Transform Spectrometer, which allows a spectral resolution in acquaintance with the precision need, and, has smaller dimensions than classical spectrometers compatible with micro satellite platform.
An on ground breadboard is being built, its aim is to experiment the instrumental concept and more precisely to test this concept for CO2 measurement. The spectrometer is centred on a CO2 absorption band at 1573 nm (6357 cm-1). Spectra are recorded on a spectral window 5.5 nm (22.5 cm-1) wide with a spectral resolution of 0.04 nm (0.15 cm-1). A year long CO2 measurement campaign is planned to examine the feasibility of the atmospheric column concentration measurement.
Distributions of nitric acid in the troposphere and the stratosphere derived from satellite measurements in the infrared
Université Libre de Bruxelles
Reactive nitrogen compounds play an essential role in processes that control the ozone abundance in the low atmosphere. There remains, however, significant lack of data regarding both the distributions of some nitrogen oxides (NO, HNO3, PAN) in the troposphere and basically of all NOy compounds at higher altitudes (upper troposphere and stratosphere).
In this work, we analyze the distributions of HNO3 retrieved from the solar occultation measurements collected by the Atmospheric Chemistry Experiment (ACE) on the Canadian Scisat satellite in 2004 and from the nadir-looking Interferometric Monitor of Greenhouse Gases (IMG), which operated onboard the ADEOS platform between 1996 and 1997.
The HNO3 distributions obtained by ACE in the upper troposphere and the stratosphere are presented and discussed by comparison with the distributions of other nitrogen oxides, obtained by the same instrument.
HNO3 abundances provided by the IMG measurements contain some vertical information and permit us to derive, for the first time, global distributions of HNO3 in the troposphere and the stratosphere. These results for the troposphere are discussed by comparison with the distribution of NO2 obtained by the Global Ozone Monitoring Experiment (GOME) instrument for the same period in April 1997.
Validation of NINFA aerosol optical thickness in the Po-Valley with AERONET sunphotometer measurements
In this poster, the ability of the air quality modelling system NINFA to calculate aerosol optical thickness (AOT) in the Po-Valley is evaluated. NINFA (North Italian Network to Forecast Aerosol pollution) forecasts daily NO2, O3, SO2 and aerosol (PM10/PM25) concentrations (horizontal resolution 10 km) for the Northern part of Italy (www.arpa.emr.it/smr/pagine/ambiente/nordItalia/).
It uses the three dimensional chemical transport model Chimere (http://euler.lmd.polytechnique.fr/chimere), driven by the meteorological model LAMI (http://cosmomodel.
cscs.ch/public/various/operational/arpa/operationalAppsARPA.htm). Chimere describes the most important phenomena affecting atmospheric pollutants: emission, diffusion, transport, chemical reactions, depositions. Furthermore it contains an aerosol module describing aerosol size distribution, -dynamic processes and – chemistry.
AOT is calculated from the NINFA output considering the aerosol extinction coefficient (Rayleigh and Mie theory) and the vertical mass distribution. It is compared with AERONET sunphotometer measurements at Modena and Venice. Results show good agreement.
As soon as satellite data with a higher spatial- and temporal resolution are available, NINFA AOT will be also validated with satellite data. If the validation is succesfull, a lookup table, generated by NINFA, can be used to retrieve directly NO2, SO2, O3 and aerosol concentrations /profiles from satellite AOT measurements.
Search of solar induced effects on the ozone layer
Istituto di Scienze dell'Atmosfera e del Clima (CNR)
Present knowledge related to solar induced effects on the ozone layer variability is discussed. The following four main sources of variability are considered:
i. the electromagnetic solar radiation;
ii. the solar wind, the electron precipitations and auroral activity;
iii. the changing galactic cosmic ray incoming;
iv. the transient effects induced by the arrival of energetic solar particles.
In particular, examples of the mesospheric depletions during energetic solar particle events are illustrated.
How good are simulated water vapour distributions in the upper-troposphere/lower-stratosphere region?
Department of Meteorology, University of Reading
This project involves comparison of UTLS water vapour predictions from the ECMWF forecasting system with independent sources of data, especially satellite data. The aim is to identify shortcomings in the model and to devise ways of correcting them. The poster will present the results of preliminary comparisons.
The potential of SCIAMACHY hydroxyl airglow emissions to derive atomic oxygen and hydrogen in the mesopause region
The energy budget of the upper mesosphere - lower thermosphere (UMLT) region is significantly determined by atomic oxygen and atomic hydrogen. Both species are very difficult to measure directly, and global datasets are rare. One possibility to derive these species is via the measurement of vibrationally excited OH.
Vibratationally excited OH is produced in the O3 + H -> OH + O2 reaction, which is the most important loss mechanism of ozone during nighttime. By applying a detailed non-LTE model of OH considering the various production and loss mechanisms, the chemical heating can be directly derived. If ozone abundance is measured simultaneously, one can also derive atomic hydrogen denisities, and, in addition, atomic oxygen densities.
The ENVISAT satellite gives a unique possibility to derive all of these quantities by the combination of SCIAMACHY and GOMOS data. Nighttime limb measurements of SCIAMACHY extend from 75 to 100 km altitude. They cover the UV, visible and near infrared region. In this study, data in the wavelength range from 1 µm to 1,75 µm is used. GOMOS star occultation measurements provide nighttime ozone abundance in the mesosphere and lower thermosphere.
Lagrangian Diagnostics of Tropical Cirrus
Rosenstiel School of Marine & Atmospheric Science
Cirrus clouds associated with tropical deep convection play an important role in regulating Earth's climate by influencing the radiative and moisture budgets of the upper troposphere. In this study, we seek to better understand the evolution of such clouds by creating a Lagrangian data base of convective systems. Specifically, we are tracking 250km x 250km water vapor spatial patterns in hourly MTSAT imagery using cross-correlations. The end product of our tracking algorithm is a Lagrangian data base of cloud trajectories and associated cloud properties and sea surface temperatures documenting the origin, evolution, and decay of cloud systems. We are obtaining cloud properties (such as optical thickness, effective radius, and ice water path) from the VISST/SIST algorithm developed by the NASA Langley Cloud and Radiation Research Group (Minnis Group) and microwave sea surface temperatures from Remote Sensing Systems. Our poster details the methodology behind our automated cloud tracking algorithm, and presents a preliminary statistical summary of the evolution of cirrus properties as a function of sea surface temperature and their effects upon the downstream upper tropospheric humidity.
Retrieval of Carbon Monoxide from MIPAS measurements
The Michelson Interferometer for Passive Atmospheric Sounding (MIPAS), is operating on Envisat since March 2002, measuring high-resolution atmospheric limb emission spectra in the interval from 685 to 2410 cm^-1 with a resolution of 0.025 cm^-1. For each orbit MIPAS performs 75 limb scans, each made of 17 spectra, 14 orbits per day.
In order to manage this amount of data MIPAS/Envisat Payload Data Segment Level-2 analysis is focused on the retrieval of profiles of pressure, temperature and volume mixing ratio of six target species (H2O, O3, HNO3, CH4, N2O and NO2). Nevertheless, spectra contain signatures of various other species. In particular at IFAC the Optimised Retrieval Model code (ORM) has been used to perform the retrieval of profiles of CFC-11, CFC-12, ClONO2 , N2O5 and CO.
In this paper we present the results of CO profile retrieval and on tropospheric CO monthly concentration mean maps.
Validation of OMI total ozone using ground-based brewer observations
Laboratory of Atmospheric Physics
Near-to-real time as well as “archive quality” Brewer total ozone observations, which are performed with well
maintained and calibrated instruments over the Northern Hemisphere have been used for the validation of the total
ozone column product of the Ozone Monitoring Instrument (OMI) aboard the NASA EOS-Aura satellite. During the
commissioning phase of OMI, the near-to-real time ground-based data, which are submitted to the WMO Northern
Hemisphere Ozone Mapping Centre within few hours after observation, have been employed to check the behaviour of
the OMI instrument as a function of measuring geometry. In addition the near-to-real time ground based data are also
used as an early warning tool for the detection of possible problems during the operation of OMI. Archived groundbased
data have been used to validate more than one year of OMI-TOMS and OMI-DOAS total ozone measurements.
The comparisons show an agreement of better than 1% for the OMI-TOMS measurements and better than 2% for OMIDOAS.
Retrieval of AOD from ground based Brewer spectrophotometer measurements in Rome
University of Rome La Sapienza
Several studies have pointed out the important role of aerosols in the Earth’s atmosphere and their impact on global climate. Reliable long time series of aerosol optical properties are still not available and there is no satisfactory worldwide spatial coverage. Here a methodology to retrieve aerosol optical depth (AOD) from Brewer direct sun measurements in the UV and visible regions is presented, together with the preliminary results for its application to the Brewer station of Rome (#067).
Homogeneity of the Vaisala radiosonde RH record
Finnish Meteorological Institute
The results of this study suggest that it is inevitable that the past changes in instrument performance will affect the homogeneity of the radiosonde humidity (RH) record. The effects of developing measurement technology are demonstrated best in the comparison of radiosonde data and ERA-40 time-series. In this comparison, climatologies for the whole record, and its sub-periods, have been compared and the effects due to the changes in sonde generations was estimated.
In addition, the differences between various Humicap generations (RS80-A, RS90, RS92) and FN-sondes, were further assessed by using the LAUTLOS-WAVVAP radiosonde hygrometer comparison results.
The performance of the RS80-A in the upper troposphere has been debated for some years, and several correction algorithms have been presented for correction of these errors (Leiterer et al. , 2000, Miloshevich et al., 2001). The main problems with Humicap measurement technology are related to the temperature-dependent dry bias and time lag, and in addition, the effects of chemical contamination (in the RS80-A). In this study, the impact of these corrections were evaluated. The corected RH climatologies over Finland revealed that the correction of the past two decades of radiosonde RH data will have a drastic effect to the trend.
Exploiting synergies of global land cover products for carbon cycle modeling
This poster addresses the user community of global land cover products. The overall objective is to present a straight forward method that merges existing products into a desired classification legend. This process follows the idea of convergence of evidence and generates a ‘best-estimate’ data set using fuzzy agreement. The method is applied to develop a new joint 1 km global land cover product (SYNMAP) with improved characteristics for land cover parameterization of the carbon cycle models that reduces land cover uncertainties in carbon budget calculations.
The overall advantage of the SYNMAP legend is that all classes are properly defined in terms of plant functional type mixtures, which can be remotely sensed and include the definitions of leaf type and longevity for each class with a tree component.
Corroboration of SYNMAP against GLCC, GLC2000 and MODIS land cover products reveals improved agreement of SYNMAP with all other land cover products and therefore indicates the successful exploration of synergies between the different products. SYNMAP is available on request from Martin Jung.
Spatial data-base of springs with the Dlubnia drainage basin as an example.
Department of Geoinformatics and Applied Computer Science, Faculty of Geology, Geophysics and Environmental Protection, AGH University of Science and Technology
For the past few years, creation a massive database of springs was a frequent issue raised during several scientific conferences in Poland. Up to now the database of that kind does not exist. The poster presented shows an example of the spatial data-base used for storage and presentation of geographical and hydrogeological characteristics of springs, with the springs of the Dlubnia drainage basin (S. Poland) as an example. The data-base was created using GeoMedia Professional and Microsoft Access packages. The information stored covers the following characteristics of springs: geographical position, aquifer description, water discharge, physical and chemical parameters of water.
The created database is to be used by Department of Hydrology, Jagiellonian University, Cracow, for gathering, storing, updating and modifying data obtained during field research. Application uses topographic maps along with ortophotomaps to visualize the researched area and the springs themselves.
While created database covers only a fraction of springs in Poland, it may serve as an example for further development and a model of using relational database along with GIS technique in hydrological database applications.
Satellite estimation of biophysical parameters for ecological models
University of Wales Swansea
The aim of this study is to establish a high-level framework for the development of future satellite sensors linking ecological models requirements and satellite capabilities.
The study focuses on multi-spectral sensors with multiple viewing angles and the parameters considered are: cover fraction, leaf area index (LAI), effective fraction of absorbed photosynthetically active radiation (fAPAR), leaf chlorophyl content and aerosol optical thickness. Two widely used ecological models - JULES and BiomeBGC - are evaluated over three boreal coniferous forests.
The soil–vegetation–atmosphere radiative transfer is simulated using a combination of the PROSPECT, FLIGHT 5.5 and 6S models. These models are inverted by means of a Look-Up-Table (LUT) approach. The advantage of the LUT technique is that it allows inversion of any model with a minimum of simplifying assumptions A second advantage is its low computer resources requirement at run-time as the complex calculations are carried out once in advance.
Top-of-the-atmosphere (TOA) reflectances with different levels of noise are used as “measured reflectances” in the retrieval of parameters from this LUT. The retrieved parameters are then used to evaluate the sensitivity of ecological models in terms of net primary production (NPP) estimated. Optimal spectral and directional sampling configurations are analysed as well as the effect of different levels of radiometric noise.
Statistical Learning Methods for Improving the Efficiency in Landscape Image Clustering and Classification Problems
TUBITAK - BILTEN
Remote sensing techniques are vital for early detection of several problems such as natural disasters, ecological problems and collecting information necessary for finding optimum solutions to those problems. Remotely sensed information has also important uses in predicting the future risks, urban planning, communication.
Recent developments in remote sensing instrumentation offered a challenge to the mathematical and statistical methods to process the acquired information. Classification of satellite images in the context of land cover classification is the main concern of this study. Land cover classification can be performed by statistical learning methods like additive models, decision trees, neural networks, k-means methods which are already popular in unsupervised classification and clustering of image scene inverse problems. Due to the degradation and corruption of satellite images, the classification performance is limited both by the accuracy of clustering and by the extent of the classification. In this study, we are concerned with understanding the performance of the available unsupervised methods with k-means, supervised methods with Gaussian maximum likelihood which are very popular methods in land cover classification. A broader approach to the classification problem based on finding the optimal discriminants from a larger range of functions is considered also in this work. A novel method based on threshold decomposition and Boolean discriminant functions is developed as an implementable application of this approach. All methods are applied to BILSAT and Landsat satellite images using MATLAB software.
Estimation of crop biophysical parameters as tool for precision agriculture
Centro de Investigacion y Tecnologia Agroalimentaria de Aragon
Remote sensing constitutes an important tool that could help to precision farming techniques that are being applied in the last years. The purpose of this study is providing information about the crop development that could be used as a source of information in precision agriculture systems. Leaf Area Index (LAI) was estimated on two different types of canopies, vineyard and corn, applying radiative transfer models (RTM) on Quickbird imagery. Modelization of reflectance was carried out using Markov Chain Canopy Reflectance Model (MCCRM) and a specific 3D model for corn and rowMCCRM on vienyards. Iterative optimization and scaling up were applied to retrieve LAI from reflectance showing acceptable accuraccy in estimations (RMSE=0.50 for corn canopies and RMSE=0.38 for vineyards)
Soil parameters (electrical conductivity, presence of different minerals...) and crop yield were also collected in situ and georeferenced. This field information was plotted against LAI maps generated by means of RTM showing that growing patterns described in satellite imagery are related with final crop yield and soil charateristics that are present in the field. The integration of generated spatial information (LAI, soil and yield maps) in a GIS could help in taking decisions within precision farming systems.
An automatic method for operational calibration of AVHRR reflective time series data
The Advanced Very High Resolution Radiometer (AVHRR) data record acquired by instruments on the NOAA polar orbiting spacecraft series comprises the longest existing daily remote sensing dataset (1979-2005). Inter- and intra-satellite factors affect the usability of these data for the generation of consistent time series. The temporal variation of the instrument sensitivity in the short-wave reflective channels needs to be addressed to ensure that derived trends are not sensor artifacts.
We describe a new method using the Multivariate Alteration Detection (MAD) algorithm to automatically select invariant features from multiple image pairs that are then compared to assess change in instrument sensitivity. This method requires no regional knowledge and is globally applicable. The derived calibration time series is shown to remove long term trends from Pseudo Invariant Features (PIFs) located in central Australia. The resulting MAD-based calibration has a root mean squared error of ~5-6% for both channel 1 and 2 and is in alignment with other approaches, and is preferable due to its operational nature.
Structural quantification of vegetative canopies based on close-range remote sensing: from 2-D to 3-D.
Katholieke Universiteit Leuven
Rapid, reliable and objective estimations of leaf area index (LAI), defined as one half the total intercepting area per unit ground surface area, are essential for numerous studies of atmosphere-vegetation interaction, as LAI is very often a critical structural parameter in process-based models of canopy response to global
environmental change. The usefulness of indirect optical LAI measurements by means of hemispherical canopy photography has already been demonstrated in that context. LAI is then calculated by gap fraction inversion. The interpretation of gap fraction in terms of LAI from optical in situ methods is based on light extinction models, which link LAI and canopy structure to the penetration of solar radiation through the canopy. This approach however is limited to 2-D since the records used are 2-D projections of the vegetation canopy,and consequently the canopy profile (z-dimension) is added based on model assumptions. The advances in terrestrial laser-scanning of vegetation during the last few years are quite promising and resulted in a variety of vegetation structure reconstructions that are based on the evaluation
of 3-D point clouds, which allow for adding the actual z-dimension for the LAI estimations in vegetation canopies.
Multiangular Earth observation: validation of information for improved terrestrial carbon cycle understanding
CARTEL, University of Sherbrooke
In the context of growing concerns about global warming and its potential consequences on humanity, reliable information about the sources and sinks of greenhouse gases are increasingly needed. CO2 has the most significant contribution to the greenhouse effect; its concentration in the atmosphere has increased considerably since the industrial revolution, in part due to fossil fuel burning. Part of these emissions has been absorbed by the Worlds oceans and trees, which act has sinks, but in the case of forests can also act has sources depending on management regime and natural events. Better understanding of this terrestrial component of the carbon cycle is a crucial element of our future ability to modify the global carbon cycle. This research aims at validating an Earth observation method developed to characterize forest structure at large scales. This information is of great importance for carbon cycle modeling, and knowledge on its accuracy is essential. The main objective is the comparison of the output of model inversion on MISR data with exhaustive field measurements taken simultaneously with the satellite passage. It will be carried out in a savanna environment of north-east Argentina; savannas being highly used and often disturbed, and consequential for the global carbon exchange flux.
Remote sensing derived data for forest management modeling
Max Planck Institute for Biogeochemistry
While globally pristine forests decline in area, managed forests expand. Science of managed forests and the goods and services they deliver - such as timber, bioenergy supply and carbon sequestration - will become more important in the future.
The development of forest ecosystems under different management scenarios can be projected into the future with the help of forest sector models. Crucial information is data on forest area and biomass distribution which is in general provided by forest inventories. However, this information is not geographically explicit. At the current state remote sensing is capable to deliver reliable estimates of forest area. From the comparison of satellite time series regional management history can be observed which facilitates the geographical allocation of forest age class structure. To detect the current forest biomass distribution and other production related parameters, a combination with ground based information from sample plots is needed. Large scale information on forest distribution, biomass and productivity can provide both, input and validation data for forest management models.
We want to use remote sensing products to make inventory information more spatially explicit and improve forestry model estimations of trends and potentials associated with forest management and its services.
Preserving Switzerland's natural heritage: Remote Sensing for the protection and conservation of Swiss dry meadows and pastures.
Dry grasslands are amongst the most species rich habitats of Switzerland. They are the result of centuries of sustainable land use by man and their distinctive compositional and structural characteristics depend greatly on climate, topography and the cultural history of each area. In addition, dry grasslands are very important for nature conservation since 40% of plant and in some cases over 50% of the animal species present on these dry grasslands are included in the red lists and are classified as endangered or threatened. However, these species-rich grasslands are endangered. It is estimated that over the past 60 years their area has declined by about 90% mainly due to intensification of agriculture.
In this poster we present a remote sensing approach for improving the identification and assisting with the monitoring of dry grasslands in Switzerland. We develop and apply a methodology using information obtained from remote sensing sensors operating at different spectral, spatial and temporal scales. We start at the plot-field scale using ASD field-spectroradiometer recordings then move to a regional scale with Hymap and Hyperion data and eventually come to the landscape scale with the use of Landsat TM/ETM+ data.
The general outline and specific results of our approach are presented.
Assimilation of meteorological and satellite data in snowmelt runoff models
ENVEO IT GmbH
In mountainous terrain, topographically induced variability of meteorological parameters governing snowmelt is complex, making spatially distributed estimates an essential requirement in alpine hydrology. Therefore preprocessing modules have been developed, accomplishing temporal and spatial assimilation of either meteorological point- (measurements) or raster-data (model forecasts). If the database comprises different time scales temporal integration is applied, for spatial interpolation the IDW-method is used. Besides meteorological data, snowmelt runoff modelling requires spatially detailed information on snow cover, which is derived from satellite observations using automated classification procedures. In order to minimize the gaps between satellite acquisitions, snow cover information obtained from different sensors (ASAR, MERIS, MODIS) are used in common. To correct for sensor specific differences in the obtained snow cover, either caused by different imaging geometry or target interaction mechanisms, a simple statistical approach is used. Daily values of snow extent on days without satellite coverage are obtained by applying a snow depletion model based on meteorological data. The use of the assimilated datasets is demonstrated in a semi-distributed snowmelt runoff model, which is operated on a daily basis in two high Alpine catchments.
Modeling the relationship between directional and hemispherical thermal emission
António José Rocha
Upwelling long wave radiation (UPLW) over land is required to close the surface energy budget and is an important input parameter to many land surface process models. Over large areas, UPLW must be determined indirectly since it is not possible to directly measure hemispherical parameters at high temporal and spatial scales using remote sensing techniques. Land Surface Temperature (LST), a common operational satellite product, is closely related to UPLW, although it has limitations since it varies with sun-view geometry. One way to estimate hemispherical UPLW from directional LST retrievals is to use a physically-based model of radiation angular anisotropy, such as the Modified Geometric Projection (MGP) model. MGP is based on a widely-used Geometric Optics model, and it’s computationally fast such that it can be used with global data sets.
In this study, the determination of the view angles at which radiance is most closely correlated with hemispherical UPLW is investigated. A sensitivity analysis of the MGP to the surface parameters that most strongly affect this relationship is also performed.
Current results were tested comparing the UPLW MGP estimates with tower-based pyrogeometer data over a savannah site in southern Africa collected during SAFARI 2000 experiment.
Crop Drought Stress Monitoring by Remote Sensing
Institute of Surveying, Remote Sensing and Land Information, University of Natural Resources and Applied Life Sciences, Vienna
The aim of this project is to adapt and develop remote sensing based methods of detection and monitoring of drought stress of agricultural crops (wheat and maize) exploiting the potentials of optical remote sensing and the synergetic effects of various sensor types offering different levels of spatial, spectral and temporal resolution. To this end, physical vegetation canopy models describing the relationship between drought stress level and reflectance characteristics of the plants are being adapted and improved. The canopy reflectance models are usually divided into two parts: modules to calculate the reflectance at leaf level (e.g. PROSPECT) and routines to calculate the reflectance of the whole canopy (e.g. SAIL). Drought stress may influence plant reflectance at leaf and at canopy level. Biophysical crop parameters indicating drought stress such as chlorophyll content and the leaf area index (LAI) can then be derived from reflectance measurements by inversion of the reflectance models using artificial neural networks or a look-up table approach. The applicability of data assimilation techniques to the problem of crop drought stress monitoring by remote sensing is discussed.
Ephemeral Water Resources Assessment in Arid Lands
Gaia Vaglio Laurin
The influence of temperature and precipitation climate regimes on vegetation dynamics: A satellite bioclimatology case study
Department of Geography, University of Cambridge
AVHRR-derived NDVI data are widely used in global-change research, yet relationships between the NDVI and ecoclimatological variables are not fully understood. This study attempts to better define these relationships by modelling climate-driven vegetation dynamics through a multivariate, spatio-temporal analysis of satellite-derived NDVI data and ground-based meteorological data in the U.S. Great Plains. Monthly maximum value composites of NDVI data (8-km resolution) and monthly temperature and precipitation records from 305 stations were collected from 1982 to 2001. Analyses involving deseasonalized datasets supported temperature as the dominant climate regime, demonstrating a higher average NDVI/temperature correlation (r = 0.73) than the NDVI/precipitation relationship (r = 0.38). The PCA also supported temperature as the dominant climate regime, accounting for 43.1% of the variance in the spatial distribution of NDVI. Cluster analysis was used to develop a climate regionalization scheme based primarily on temperature, and NDVI characteristics of each sub-region were compared. The statistical modeling methods applied were useful in capturing and characterizing the seasonal response of NDVI to climate variability. In the context of global climate change, findings from this satellite bioclimatology study emphasize the influence of temperature and precipitation variability over vegetation cover in the Great Plains region.
Remote Sensing Image Classification Method Based on Geostatistics and ANN
China Institute of Water Resources and Hydropower
Texture is the key character of remote sensing image.In this paper, the image texture is extracted by means of semivariogram. On the base of this, the study adopts the back propagation artificial neural network method to classify Combining spectral feature with many sort of textures. Then the classification results are compared with those gained by maximum likelihood method and the results of the study proved that the way that combining spectral features and textural measures based on the geostatistics and NN theory to the classification of the remote sensing image may improve the accuracy of image classification.
Fire severity mapping using Landsat 5 - TM, Envisat - MERIS and Terra - MODIS post-fire images
Spanish Institute for Agricultural Research (INIA) - Remote Sensing Laboratory
This analysis concerns an estimation of burned area and fire severity levels in an area affected by a large wildfire that took place in the South of Spain (Huelva-Sevilla) in July 2004. Fire severity is defined in this work as the impact caused in vegetation by a fire. The objective was to find an efficient method for quick fire severity mapping based on remote sensing techniques that can be useful fos post-fire forest management. Several methods for image analysis (Linear Spectral Unmixing, Matched Filtering, and Normalized Burn Ratio Index) were applied to post-fire Landsat 5-TM, Envisat-MERIS and Terra-MODIS images. Maps depicting fire severity of three levels of an acceptable reliability were obtained using a small amount of field data and following a simple method of processing. Linear Spectral Unmixing produced the best classifications for MERIS and MODIS images, while the Matched Filtering technique produced the most accurate classification for the TM image. These preliminary results show that short-term severity maps can be obtained by means of high to medium resolution post-fire remote sensing data, in order to evaluate the situation after a forest fire and plan forest restoration works.
Assessment of time-dependent biases in the MODIS Land Surface Temperature (MOD11_L2) product
New University of Lisbon
Land surface temperature (LST) is a key land parameter to estimate the energy and hydrologic state of the Earth's surface. Over large areas, this parameter is typically retrieved from moderate resolution sensors (e.g., AVHRR, MODIS, AATSR) on polar-orbiting satellites (e.g., POES, EOS Terra/Aqua, ENVISAT). However, these wide-field-of-view (~2000-3000 km wide swath) sensors can observe land targets at very different local times (e.g., hours apart) within a single sub-second scan. Since instantaneous LST depends in part on environmental and sampling variables that change predictably with time (e.g., cumulative solar heating, atmospheric state, sun-view observation geometry), systematic measurement biases may exist based solely on pixel position within a swath. The goal of this study is to determine if statistically significant temporal biases exist within swath LST data, and if so, to evaluate their magnitude as a function of latitude, time of day and year, land cover type, view geometry and other ground and observational parameters. We use data from the MODerate-resolution Imaging Spectroradiometer (MODIS) LST product, MOD11_L2 swath scenes, 1 km spatial resolution (at nadir), in our analysis. We focus our study on the African continent and the year 2001. We will discuss potential errors in our approach, and will conclude by proposing a method to correct within-swath temporal biases.
Identification of Sea Ice Catchment and Snow Distribution from Remote Sensing Data
Centre for Earth Observation Science
The ocean-sea ice-atmosphere (OSA) interface, is an important region for mass, gas and energy transfer in the Arctic marine cryosphere. Snow regulates the growth and decay of sea ice through its control on conductive and radiative fluxes across the OSA interface. Through these processes snow moderates the global climate and is a major controlling factor in the ecology of the Arctic system from algal production in the ocean to habitat selection of apex predators such as polar bears and whales. The pattern of snow distribution is primarily controlled by the ice surface topography (Iacozza and Barber, 1999; Eicken et al., 1994; Adolphs, 1999; Jeffries et al., 1995). These variations in the snow distribution patterns for various ice types have significant influences on understanding the physical-biological coupling. The primary objective of this research is to investigate the relationship between sea ice surface roughness and snow distribution for different classes of first-year sea ice at the satellite scale. This will be accomplished by (i) characterizing ice roughness for fast and mobile ice in Franklin Bay using EM induction data, (ii) relate the variability at the local scale to that for satellite imagery using ASAR data, and (iii) linking ice roughness to snow catchment at the satellite scale.
An Optimization Approach to Modelling Sea Ice Dynamics
University of Washington
A new model for the dynamics of sea ice is explored. The pressure field, instead of being derived from a local rheology, as in most existing models, is computed from a global optimization problem. Here the pressure is seen as emerging not from an equation of state but as a Lagrange multiplier that enforces the ice's resistance to compression while allowing divergence. The resulting variational problem is solved by minimizing the pressure globally throughout the domain, constrained by the equations of motion along with the natural limits on ice concentration. This formulation has an attractive mathematical elegance while being physically motivated. Moreover, it leads to an analytic formulation that is also easily implemented in a numerical code, which exhibits marked stability and is suited to capturing discontinuities.
The theory is initially tested in a one-dimensional model in Lagrangian mass coordinates. The model results are compared to an exact analytic solution for a simple test case, as well as to a particle-resolving model. After casting the model in Eulerian coordinates, a finite ice strength is introduced, permitting the important process of ice yielding to be captured.
Geometric changes of Austfonna Ice Cap, Svalbard
University of Oslo
Austfonna, 8200 km2, is one of the largest arctic ice caps outside of Greenland. Apart from a few studies, little is known about its response to recent climate change. As a part of the ESA collaborated campaign for calibration and validation of the future CryoSat, the University of Oslo and the Norwegian Polar Institute have set up an extensive program on Austfonna which involves annual fieldwork of key glacier parameters. Ground truth data are compared with simultaneous CryoSat simulation flights in order to correlate physical properties of the snowpack with penetration depth and volume scattering of the CryoSat radar signal.
Elevation changes on the ice cap are estimated from annual GPS/GPR-profiles, airborne radar/laser overflights and satellite altimeters such as ICESat, ERS/ENVISAT and future CryoSat. Some of these measurements willl be combined with SAR interferometry from ERS/Envisat in order to produce a more accurate DEM. In terms of geometric changes, the dynamic regime of the ice cap also needs to be considered. Surface velocities are measured on ground by repeated GPS measurements, and velocity fields can be identified by differential InSAR.
Wind-blown snow interactions with a rift in the Ross Ice Shelf, Antarctica
Lamont-Doherty Earth Observatory / Columbia University
The Nascent Rift in the Ross Ice Shelf, Antarctica has persisted for several years, with a distinctive profile of ice mélange (a mixture of windblown snow, marine ice, and pieces of ice talus broken from the ice shelf) filling it at a depth of roughly 30m below the surrounding shelf surface. Modeling results obtained using the Piektuk-Tuvaq blowing snow model (Dery & Tremblay, 2004) suggest that the contribution of blowing snow to this mixture is significant. Windspeeds sufficient to entrain snow into suspension occur during 20% of the year (73 days). Coincidentally, 73 days of 25m/s wind (at the 10m height) could fill the 100m wide by 30m deep rift with snow, while 16 years of 7.5 m/s winds are required to deposit this same volume, due to the non-linearity of the suspended snow mass transport Qt with windspeed. The saltating horizontal mass flux of snow at low windspeeds is significant relative to the suspended blowing snow mass flux. As the saltation threshold for snow is substantially lower than the threshold for suspension, the annual mass flux of snow via this mechanism is important, due to the greater amount of time when the region experiences such low windspeeds.