Name COASTALT-GNSS
Title Development of Radar Altimetry Data Processing in the Oceanic Coastal Zone
Thematic Area Ocean
Cost < 100 K
Action Line Novel Algorithms and Products
Status Completed in 2009
Missions ENVISAT
Sensors RADAR ALTIMETER 2, Microwave Radiometer (MWR)
Objectives This project will contribute to the definition, specification, development and validation of a new pulse-limited coastal zone radar altimetry product which is intended to become operationally processed, including the reprocessing of the ESA Radar Altimetry archive (ERS-1, ERS-2, ENVISAT). This activity is focused on two main aspects: 1. Improvement of the wet tropospheric correction. The wet tropospheric correction is one of the major error sources in coastal altimetry. The most accurate method to derive the correction, from Microwave Radiometer (MWR) measurements, becomes unusable at distances about 50 km from the coast due to the large MWR footprint. This study is focused on the development of an innovative method for computing the wet tropospheric correction for altimetry measurements in the coastal regions from GNSS‐derived tropospheric delays, the so‐called GPD (GNSS‐derived Path Delay) approach. In addition, a backup approach is also proposed, the Dynamically-Linked Method (DLM), to be used whenever GPD is not available. a) DLM This is a simple and easy to implement procedure that requires only data from the altimeter Geophysical Data Records (GDR) with optional information from a distance-to-land global grid and can be implemented globally for any satellite. DLM takes full advantage of the two types of wet tropospheric correction that are present on the GDR: the MWR-derived correction and a large-scale atmospheric reanalysis numeric weather model (NWM) derived correction such as that from the European Centre for Medium‐Range Weather Forecast (ECMWF). The method consists in replacing, in the coastal regions, the invalid MWR-derived correction by the ECMWF correction, somehow dynamically linked to the closest points with valid MWR field, to warrant continuity. This approach is quite different from the use of the model correction everywhere since it significantly reduces the well known large-scale errors of most NWM wet tropospheric corrections. Studies previously conducted by Fernandes et al. (2003) on ERS data, show that this simple approach leads to a data recovery of 80 to 90% of the invalid measurements in the coastal regions, does not introduce discontinuities in the correction and can be used to generate coastal products in an operational processing scheme. Tests are being performed for Envisat, using the two wet tropospheric correction models provided in the GDR: ECMWF and NCEP (National Center for Environmental Prediction, USA). On average, for Envisat cycle 54, 92% of the invalid coastal points are adjusted. For the unadjusted measurements the output field shall be the model field corrected for the mean bias between the MWR and each model field. b) GPD The method is based on GNSS‐derived zenith wet delays (ZWD) determined at a network of coastal stations and offshore platforms or buoys equipped with dual‐frequency GNSS receivers, further combined with valid MWR measurements and ZWD values from a NWM (e.g. ECMWF). A methodology for computing the wet tropospheric correction at each altimeter point with invalid MWR measurement is implemented by using a linear space‐time objective analysis technique. The statistical technique interpolates the wet correction measurements at each altimeter ground‐track point with invalid MWR measurements from the nearby (in space and time) valid MWR-, ECMWF‐ and GNSS‐derived independent measurements, and takes into account the accuracy of each data set. An added value of this method is that it provides a quantification of the error associated with each estimated ZWD value. Results obtained so far for the whole ENVISAT data series, along the west Iberian and Mediterranean regions, show that the GPD estimates are highly dependent on the spatial and temporal distribution of the three data types used. A considerable number of configurations can be found, which allow the estimation of the wet delays within 1 cm error: points at distances < 50 km from a GNSS station, points for which there are valid MWR measurements within a distance < 50 km or passes with an associated measurement time very close in time to one NWM grid. Comparisons of GPD output with ZWD determined by other independent methods are also on going. 2. Geophysical validation and performance assessment in a small region in Portuguese Coastal Zone. This activity will assess the performance of the coastal altimetry product to reproduce the variability observed in the ground-based measurements in the study region. Namely, to obtain sufficient in-situ information from coastal tide/wave gauges and other instrumentation and numerical models to enable an adequate comparison of existing and improved altimetry data sets near the coast and thereby an evaluation of the improvement in performance. The main beneficiaries of this development are scientists interested in observing coastal zone oceanographic processes.
1. Improvement of the wet tropospheric correction. The wet tropospheric correction is one of the major error sources in coastal altimetry. The most accurate method to derive the correction, from Microwave Radiometer (MWR) measurements, becomes unusable at distances about 50 km from the coast due to the large MWR footprint. This study is focused on the development of an innovative method for computing the wet tropospheric correction for altimetry measurements in the coastal regions from GNSS‐derived tropospheric delays, the so‐called GPD (GNSS‐derived Path Delay) approach. In addition, a backup approach is also proposed, the Dynamically-Linked Method (DLM), to be used whenever GPD is not available. a) DLM This is a simple and easy to implement procedure that requires only data from the altimeter Geophysical Data Records (GDR) with optional information from a distance-to-land global grid and can be implemented globally for any satellite. DLM takes full advantage of the two types of wet tropospheric correction that are present on the GDR: the MWR-derived correction and a large-scale atmospheric reanalysis numeric weather model (NWM) derived correction such as that from the European Centre for Medium‐Range Weather Forecast (ECMWF). The method consists in replacing, in the coastal regions, the invalid MWR-derived correction by the ECMWF correction, somehow dynamically linked to the closest points with valid MWR field, to warrant continuity. This approach is quite different from the use of the model correction everywhere since it significantly reduces the well known large-scale errors of most NWM wet tropospheric corrections. Studies previously conducted by Fernandes et al. (2003) on ERS data, show that this simple approach leads to a data recovery of 80 to 90% of the invalid measurements in the coastal regions, does not introduce discontinuities in the correction and can be used to generate coastal products in an operational processing scheme. Tests are being performed for Envisat, using the two wet tropospheric correction models provided in the GDR: ECMWF and NCEP (National Center for Environmental Prediction, USA). On average, for Envisat cycle 54, 92% of the invalid coastal points are adjusted. For the unadjusted measurements the output field shall be the model field corrected for the mean bias between the MWR and each model field. b) GPD The method is based on GNSS‐derived zenith wet delays (ZWD) determined at a network of coastal stations and offshore platforms or buoys equipped with dual‐frequency GNSS receivers, further combined with valid MWR measurements and ZWD values from a NWM (e.g. ECMWF). A methodology for computing the wet tropospheric correction at each altimeter point with invalid MWR measurement is implemented by using a linear space‐time objective analysis technique. The statistical technique interpolates the wet correction measurements at each altimeter ground‐track point with invalid MWR measurements from the nearby (in space and time) valid MWR-, ECMWF‐ and GNSS‐derived independent measurements, and takes into account the accuracy of each data set. An added value of this method is that it provides a quantification of the error associated with each estimated ZWD value. Results obtained so far for the whole ENVISAT data series, along the west Iberian and Mediterranean regions, show that the GPD estimates are highly dependent on the spatial and temporal distribution of the three data types used. A considerable number of configurations can be found, which allow the estimation of the wet delays within 1 cm error: points at distances < 50 km from a GNSS station, points for which there are valid MWR measurements within a distance < 50 km or passes with an associated measurement time very close in time to one NWM grid. Comparisons of GPD output with ZWD determined by other independent methods are also on going. 2. Geophysical validation and performance assessment in a small region in Portuguese Coastal Zone.
1. Improvement of the wet tropospheric correction. The wet tropospheric correction is one of the major error sources in coastal altimetry. The most accurate method to derive the correction, from Microwave Radiometer (MWR) measurements, becomes unusable at distances about 50 km from the coast due to the large MWR footprint. This study is focused on the development of an innovative method for computing the wet tropospheric correction for altimetry measurements in the coastal regions from GNSS‐derived tropospheric delays, the so‐called GPD (GNSS‐derived Path Delay) approach. In addition, a backup approach is also proposed, the Dynamically-Linked Method (DLM), to be used whenever GPD is not available. a) DLM This is a simple and easy to implement procedure that requires only data from the altimeter Geophysical Data Records (GDR) with optional information from a distance-to-land global grid and can be implemented globally for any satellite. DLM takes full advantage of the two types of wet tropospheric correction that are present on the GDR: the MWR-derived correction and a large-scale atmospheric reanalysis numeric weather model (NWM) derived correction such as that from the European Centre for Medium‐Range Weather Forecast (ECMWF). The method consists in replacing, in the coastal regions, the invalid MWR-derived correction by the ECMWF correction, somehow dynamically linked to the closest points with valid MWR field, to warrant continuity. This approach is quite different from the use of the model correction everywhere since it significantly reduces the well known large-scale errors of most NWM wet tropospheric corrections. Studies previously conducted by Fernandes et al. (2003) on ERS data, show that this simple approach leads to a data recovery of 80 to 90% of the invalid measurements in the coastal regions, does not introduce discontinuities in the correction and can be used to generate coastal products in an operational processing scheme. Tests are being performed for Envisat, using the two wet tropospheric correction models provided in the GDR: ECMWF and NCEP (National Center for Environmental Prediction, USA). On average, for Envisat cycle 54, 92% of the invalid coastal points are adjusted. For the unadjusted measurements the output field shall be the model field corrected for the mean bias between the MWR and each model field. b) GPD The method is based on GNSS‐derived zenith wet delays (ZWD) determined at a network of coastal stations and offshore platforms or buoys equipped with dual‐frequency GNSS receivers, further combined with valid MWR measurements and ZWD values from a NWM (e.g. ECMWF). A methodology for computing the wet tropospheric correction at each altimeter point with invalid MWR measurement is implemented by using a linear space‐time objective analysis technique. The statistical technique interpolates the wet correction measurements at each altimeter ground‐track point with invalid MWR measurements from the nearby (in space and time) valid MWR-, ECMWF‐ and GNSS‐derived independent measurements, and takes into account the accuracy of each data set. An added value of this method is that it provides a quantification of the error associated with each estimated ZWD value. Results obtained so far for the whole ENVISAT data series, along the west Iberian and Mediterranean regions, show that the GPD estimates are highly dependent on the spatial and temporal distribution of the three data types used. A considerable number of configurations can be found, which allow the estimation of the wet delays within 1 cm error: points at distances < 50 km from a GNSS station, points for which there are valid MWR measurements within a distance < 50 km or passes with an associated measurement time very close in time to one NWM grid. Comparisons of GPD output with ZWD determined by other independent methods are also on going.
2. Geophysical validation and performance assessment in a small region in Portuguese Coastal Zone.
This activity will assess the performance of the coastal altimetry product to reproduce the variability observed in the ground-based measurements in the study region. Namely, to obtain sufficient in-situ information from coastal tide/wave gauges and other instrumentation and numerical models to enable an adequate comparison of existing and improved altimetry data sets near the coast and thereby an evaluation of the improvement in performance. The main beneficiaries of this development are scientists interested in observing coastal zone oceanographic processes.
Project Partners FCUP : Faculdade de Ciencias, Universidade do Porto(Prime contractor)HIDROMOD : HIDROMOD Modelacao em Engenharia Lda(Subcontractor)
Project Manager Dr. Joana Fernandes, Faculdade de Ciências, Universidade do Porto (FCUP), Departamento de Matemática Aplicada, Rua do Campo Alegre 687, 4169-007 Porto, Portugal, Tel: +351 220 402 202, Fax: +351 220 402 209,
Technical Officer Jérôme Benveniste