Recent research suggested that the digital elevation models can be considered, as an alternative to the altimetry data. However, the water prohibits the ability of monitoring the construction the waterbed, due to the loss of the returning signal. The elevation models have therefore a flat surface, which prohibits the extraction of the water level and volume variability. The goal of this thesis was to develop a tool, that uses the elevations where no flatting had occurred, and create a linear representation of the shape of the lake or reservoir. The following research question was formulated: What is the potential of applying elevation models to monitor the volume levels of lakes and reservoirs, when replacing the flat elevations with extrapolated depths? The case study comprised of two parts. First, the validation of the water level extraction was conducted for pre-selected lakes and reservoirs. The results showed an average RMSE of 4.10 [m]. When analysing the distribution of lakes and reservoirs data, the RMSE was centred at 1.27 [m] (NED and ALOS) and 3.70 [m] (SRTM). The comparison between the three considered elevation models, showed a minor improvement regarding the relative water level. At the same time, it demonstrated that the number of waterbodies that can be monitored increases when applying the developed tool. The second part of the case study aimed at providing an insight in the potential of applying the DEM for estimating the volume variability. An analysis of the USGS in-situ data compared to the results, showed that the overall RMSE decreased for each considered model. From the same analysis, the relative time-series could be determined for each model i.e., in case the minimum amount of surface area was available and in case the water level variation was significantly noticeable. The results from this work indicated the potential of the global elevation models to monitor the volume levels of lakes and reservoirs. The developed tool demonstrated that elevations can be extrapolated, which would fit a more realistic shape of the reservoir at areas where the bathymetry was flattened or not found. Accurate information for extracting the water level and assessing the volume variability of lakes and reservoirs currently depends on extensive hydrographic surveys. This work provides guidelines for an alternative method: the Linear Bathymetry for Digital Elevation Models tool.

Rivers play a crucial role in shaping landscapes and supporting ecosystems. This is demonstrated by the tiger habitats in and around Bardia National Park in West-Nepal, which rely on the Karnali River. This study contributes to a larger effort aimed at sustainably managing these tiger habitats. Monitoring the rivers in this remote area is challenging, suggesting a role for remote sensing. An exploration is presented regarding the potential of satellite synthetic aperture radar altimetry (sat-SARA) for monitoring rivers situated in diverse topographic landscapes. Focusing on the Bheri, Karnali, and Geruwa Rivers, the applicability of sat-SARA techniques for water level monitoring, multiple channel identification, and channel activation detection was evaluated. For deriving water surface heights from sat-SARA data, an empirical Gaussian retracker was used. The findings are promising. While resulting water level variations align with field observations, complementary in-situ measurements are imperative for a comprehensive evaluation. Additionally, the study reveals the potential for identifying multiple channels from sat-SARA return signals, extending to channel classification and detecting channel activation. Leveraging the labour-intensive nature of sat-SARA data processing, the technique holds great promise for monitoring rivers in remote and difficult-to-access landscapes. Therewith, this study contributes to advancing the understanding of the hydrodynamics of the Lower Karnali River and opens doors for sat-SARA applications for river monitoring in challenging terrains.

Interferometric Synthetic Aperture Radar (InSAR) and Global Navigation Satellite Systems (GNSS) are widely used to monitor the dynamic behavior of the Earth. InSAR is a geodetic technique that estimates millimeter-level relative displacement time-series in an opportunistic network of a multitude of coherent points on the Earth’s surface, in a local datum. GNSS uses ground-based instrumentation to acquire time-series data over a limited number of specific and well-defined points in a known geodetic datum.The Integrated Geodetic Reference Station (IGRS) is designed to combine these (and other) techniques into one common instrument, establishing an integrated benchmark, i.e., a GNSS antenna and two radar corner reflectors, ensuring an identical kinematic behavior. This enables a geodetic datum connection, effectively enabling the InSAR results to be represented in a common geodetic datum, instead of a free network.However, the efficacy of the IGRS has not yet been proven, i.e., a thorough analysis of the first empirical results of an IGRS network has not yet been performed.Here we show that by using three years of data from a spatio-temporal network of 29 IGRS stations in an area of 60×60 km, and 742 independent Synthetic Aperture Radar (SAR) acquisitions, we reach a high level of agreement, demonstrating that IGRS can be used to connect InSAR information products to a well-defined geodetic datum.By using an Overall Model Test with a significance level of 5% we found that 96% of the double-difference arcs in time and space, sustain the null hypothesis that both the InSAR and the GNSS results stem from the same distribution. We found that the main reason for rejecting the null hypothesis for the remaining 4% of the double-difference arcs is that the results of both the InSAR and the GNSS are affected by a leakage of signal from the functional to the stochastic model. For the InSAR observations there is inadequately modeled atmosphere leaking into the stochastic model, while for the GNSS it appears that the precision estimate of the periodically moving stations is worse than the non-periodically moving stations, which suggests that the stochastic model is influenced by the the functional model.In the end, this study proved the efficacy of the IGRS to connect different geodetic datums, and this enables the InSAR results to be integrated in a well-defined geodetic datum.

The purpose of this thesis is to do quality assessment of GNSS/IMU derived NAP heights for The Netherlands (NAP) using Fugro RILA technique and the RDNAPTRANS2018 published by Rijkswaterstaat. The use of Global Navigation Satellite System (GNSS) is growing rapidly in order to determine the position (both horizontal and vertical). However, the GNSS is only able to give the geometric height which is the position on the ellipsoid, which have a drawback that the surface of constant ellipsoidal height are not equipotential surface, and hence these heights need to be transformed into traditional height systems such as the Normaal Amsterdam Peil (NAP) used in The Netherlands. In order to derive these NAP heights from the GNSS heights, a (quasi-) geoid model along with corrector surface model is used to convert from one height to another. In this study the newly computed local quasi geoid model NLGEO2018 rather than the NLGEO2008 is adapted using RDNAPTRANS2018 along with Fugro Rail Infrastructure aLignment Acquisition (RILA) technique which makes use of GNSS/IMU to obtain ellipsoidal height. It was found upon using the older transformation procedure RDNAPTRANS2008, which makes use of NLGEO2008, a mismatch in the order of 17mm between the NAP and GNSS-derived NAP from RILA. This error is not only due to the geoid itself but also the systematic errors in the different height system. However, this new geoid NLGEO2018 along with the new transformation procedure of RDNAPTRANS2018 allows a more accurate conversion of ellipsoidal to normal heights and this study focuses on this quality assessment of the new GNSS/IMU derived NAP heights and investigate how well NAP heights be obtained using RILA technique based on the new geoid. It was found with the use of the new geoid and the new transformation model RDNAPTRANS2018, the mean height difference was reduced from 12mm to 3.6mm showing a better fit of GNSS heights to NAP heights. An error budget was also calculated, to understand the reliability of these RILA measurements with the NAP heights. This error budget includes all the uncertainties from all error sources, namely RILA derived height, geoid height and the levelled height, followed by hypothesis testing. The highest uncertainty was from GNSS/IMU measurements from RILA system followed by the levelling and then the gravimetric quasi-geoid. Finally from this analysis, areas where the terrestrial surveying can be avoided was found. It was found that near the stations, tunnels, high vegetation had the highest uncertainty due to poor GNSS reception. It was also found that level crossing also showed a constant systematic offset between the two heights potentially due to the materials of the level crossings.

Subsurface firn processes play a crucial role in ice sheet mass loss mechanisms. On Greenland surface meltwater percolates to deeper layers where porous firn retains it, directly inhibiting runoff. However, secondary effects such as the formation of impermeable ice slabs may indirectly and irreversibly accelerate runoff and with it global sea level rise. Microwave remote sensing offers opportunities to monitor these processes, but due to the simplicity of their underlying snow models retrieval methods fail over areas subject to melt and refreezing - areas where the firn's (in)ability to buffer meltwater is critical. This study presents a new forward model which given initial conditions and atmospheric forcing first solves for the firn state through a full-complexity snow model (SNOWPACK) and then simulates multifrequency brightness temperature (Tb) time series (using radiative transfer model SMRT). As part of a comprehensive sensitivity analysis three ensembles of multi-decade Tb time series (19 and 37 GHz) were modelled for the DYE-2 site in the percolation area of the Greenland Ice Sheet. Model performance based on RMSE w.r.t. independent Tb satellite observations was found to be sensitive to biases introduced in the atmospheric forcing record (with air temperature, precipitation and relative humidity controlling variance) and snow model settings (new snow grain size and albedo settings) and not to initial profile conditions. However, computed RMSEs were high (min. 17.8 K at 37V and 19.4 K at 19V) due to trends in modelled Tb consistently underestimating observed trends when taken over an accumulation season. It is shown that this can only be explained by the constant-with-time stickiness assumption used to link the snow model’s microstructure representation to the sticky hard sphere model employed for the radiative transfer scheme. A seasonal stickiness signal is made evident for the conditions at DYE-2 and linked to its yearly melt-refreeze-accumulation cycle. These results demonstrate that earlier approaches to forward model microwave satellite observations based on a constant-with-time stickiness assumption (or that lack a third snow microstructure parameter altogether) are not valid for ice sheet areas prone to melt. This study is expected to be the starting point for a more sophisticated implementation that estimates a snow layer's stickiness from microstructure information already available in the snow model. If successful it would be the first of its kind and open the door to satellite-based retrieval of subsurface firn properties and processes from areas where observations are currently lacking, greatly reducing uncertainty in ice sheet mass loss and global sea level rise projections.

Detection of the openings in the Arctic sea ice pack, or leads, allow to sample instantaneous sea surface height (SSH) and this information is crucial for quantifying the impact of sea ice melting. It is therefore important to correctly detect as many leads as possible to obtain more SSH references. This paper studies 12 different classification methods including supervised-, unsupervised machine learning methods and thresholding method, being applied to the Sentinel-3 Synthetic Aperture Radar (SAR) altimetry data collected in March/April of 2017-2020 and June/July of 2020 from areas all across the Arctic Ocean. These are compared and assessed with respect to images taken by Ocean and Land Color Instrument (OLCI), also on board Sentinel-3, ensuring a perfect temporal alignment between the two measurements. The supervised Adaptive Boosting, Artificial Neural Network and Linear Discriminant classifiers showed excellent and robust results in March/April with overall accuracies up to 91.82%. The unsupervised K-medoid classifier produced excellent results achieving up to 91.51% accuracy and it is an attractive classifier as it does not require ground truth data. The classifiers perform poorly in the summer months, as sea ice returns show more ambiguous reflections due to melting. Therefore on summer data, classifications that are solely based on waveform data from SAR altimetry is unsuitable and auxiliary information is required. Furthermore, this paper attempted to identify off-nadir leads (ONL) by adding an extra class in supervised learning methods, intending to reduce the falsely detected leads. Most classifiers failed to detect leads and did not improve their false lead rate. However, as RUS Boost classifier was able to identify 61.6% of total ONLs, this can be used to initially reject these points for more conservative lead detection.

Establishing an accurate global unified vertical reference frame (VRF) is a long-standing objective of geodesy. However, that objective has still not been achieved. One particular application where the lack of such a VRF is evident, is the improvement of hydrodynamic models by assimilating total water levels acquired by tide gauges. Indeed, to facilitate a straightforward assimilation requires that both the observed and modeled water levels refer to the same vertical datum. The required accuracy is high; it is expected to be in the order of 1centimeter. The best alternative VRF for the area of interest, the northwest European continental shelf, is the European Vertical Reference Frame 2019 (EVRF2019). The EVRF2019, however, still lacks complete coverage and the required accuracy. The key reason is that it is solely based on geopotential differences from spirit leveling/gravimetry, which are not available between benchmarks separated by large water bodies. This thesis exploits model-based hydrodynamic leveling to provide these differences. The specific objective is to assess the potential of including these data in realizing of European Vertical Reference System (EVRS).@en

As extensive efforts from consumer drone manufacturers resulted in inexpensive aircrafts that can capture high quality video imagery, drones are increasingly considered to be beneficial for scientific purposes. In the recent past, video imagery has been used to analyze waves in terms of several hydrodynamic parameters and to indicate matching coastal features. Whereas these measurements have been acquired using static cameras mounted on large poles situated at beaches, this report exploits a recently developed method using Unmanned Aerial Vehicles (UAVs) as a means to map coastal morphology. After recording aerial imagery in combination with several Ground Control Points (GCPs), several time series of georectified coastal images are compiled. Subsequently, for all of the points in a predefined grid, pixel intensities are stored throughout the time of the recording. Consequently, hydrodynamic data like wave celerity and phase are estimated, which in turn are used to invert water depths for every location in a predetermined area of interest using an algorithm called cBathy. Using reference measurements with an accuracy in the order of five centimeters, this report benchmarks the bathymetry as computed using the drone imagery by calculating the root mean squared error, the root mean squared error divided by the water depth and the mean error of three different sub areas within the area of interest. After calibrating parameters as used by the cBathy algorithm, it is shown that the best computation yielded a bathymetry with a root mean squared error of 0.37 meters for a total area of approximately 2500 square meters. It is also shown that the other datasets yield errors of approximately twice the error of the best dataset. It is shown that the total error can to a large extent be attributed to errors in the rectification part of the algorithm. The large errors and the large discrepancy between the errors make the method currently unsuitable for coastal monitoring purposes. Hence, before the UAV bathymetry mapping method is to replace traditional methods, more research should focus on standardizing the process and thereby decrease the variance between the errors of different datasets.

The topic of this thesis is Integrity Monitoring (IM), with special focus on the Singapore Satellite Positioning Reference Network (SiReNT) infrastructure. This network infrastructure supports and improves Global Navigation Satellite System (GNSS) applications, related to positioning, navigation, tracking and (deformation) monitoring. This project focuses on what IM is with regards to the SiReNT infrastructure, and how the current methodology can be improved so that its output is easier to interpret/understand for its users. This research consists of two parts, a theory study, and a data study. The theory study is aimed at developing a fundamental understanding of GNSS and integrity, and the need for an IM component, as well as what comprises IM around the globe. The data study, which includes IM data processing, is performed to develop an understanding of the stability and performance of the SiReNT infrastructure itself, and its sensitivity to several degrading factors. For this the IM displacement data of a period of up to 5 months is analysed, in combination with the atmospheric information, and the number and geometry of tracked satellites. The Trimble method, with its default thresholds, has been applied to this IM displacement data, in order to determine what percentage was supposedly reliable, and inside these thresholds. It was also investigated whether these thresholds are performing adequately, if applied, and whether they have any impact on simultaneous recorded Rover data. The latter was assumed; the IM displacement data is thought to be usable to control the Rover data. When it comes to positioning solutions, the quality parameters include, integrity, accuracy, precision, availability, reliability, continuity, capacity and redundancy. These quality parameters are affected by several degrading factors. At the moment the Trimble Pivot Platform (TPP) is used to perform the network processing, as well as the integrity monitoring of the Continuous Operating Reference Stations (CORS) of the SiReNT infrastructure (and with that the impact of several of these degrading factors). This method is an example of the so-called external monitoring method. ‘External’ as the monitoring does not take place within the receiver itself, such as in the other method available; Receiver Autonomous Integrity Monitoring (RAIM). The external monitoring method and RAIM can either be used individually and/or independently, or in some sort of combination, complementing each other. In several other countries, namely the Netherlands, Australia (Victoria), Hong Kong, and Malaysia, variants of the external monitoring method are applied in performing IM of their GNSS infrastructures. Usually this includes baseline and coordinate processing/comparison, both in real-time and through post-processing (often using the Bernese GNSS software). All of them have backup systems in place. When it comes to IM visualization, even though there are similarities, the approaches differ per country. Some only visualise as much as the status of their reference stations, others visualise the real-time positioning solution of their IM stations, or have a complete dashboard visualising a number of parameters. From this research can be concluded that the integrity (in terms of CORS’ stability and positioning solution quality) of the SiReNT infrastructure is actually quite good. Of the IM displacement data which was supposedly reliable (following from the applied method), over 99% was within the set thresholds, for all CORS. The default thresholds (therefore) are iv considered to be quite a good trade-off; they filter out most of the main disturbances (especially the ones which are assumed to be related to sudden severe fluctuations in the Ionospheric activity), while minimizing the amount of data loss. There is room for improvements however; as a start, the (overall) redundancy should be increased, making the SiReNT infrastructure more robust and increase its reliability. The method in which the IM is currently performed could be improved in several ways as well. This could for example be done by tracking more constellations/signals, constructing additional CORS and backup systems, and using multiple different hardware brands for the equipment that is used in the SiReNT infrastructure. Also, the SiReNT infrastructure should be monitored with respect to surrounding IGS stations, rather than using a CORS that is part of the SiReNT infrastructure itself as a reference. An important addition to the IM component could be the feedback from the Rover data and performance (the initialization time, and positioning solution quality). Currently made assumptions related to that do turn out to be erroneous, however. Any threshold applied to the IM displacement data cannot be used to control the Rover data. The way in which the output resulting from this IM is visualised, should be more intuitive/easy to interpret. For this a dashboard is to be developed (in design following the example of the dashboard of the Dutch Cadastre), which homepage should be visualizing several parameters, all in real-time. This would enable the (end-)user to see what is the current status with regards to all these parameters, in one view. One would however also be able to access further information related to these parameters, by selecting the one of interest. These parameters are to be the number of tracked satellites, the number of connected users, an indication of the Ionospheric delay, the predicted geometric error, and the Rover performance. The content of the information presented to the public should be kept to a minimum (only the necessary), where the content presented to the system administrators can be more elaborate. The conclusions of this research point towards the need for further research into how the applied methods/algorithms and thresholds (as well as the other available options) can be utilized in order to improve the IM procedure in Singapore. On top of that, the impact of interference should be identified and monitored. Being able to do this, will improve the (integrity of the) infrastructure and the information provided by it.

Climate change causes alterations in large scale mass transport patterns in the ocean, cryosphere and hydrology. The Gravity Recovery and Climate Experiment (GRACE) satellite mission which has been operational in the years 2002-2017 has already improved our understanding of large scale mass transport on Earth, but improvement of data quality is still needed. This will increase the quality of our current estimates of the effects of climate change on one hand and help in the validation and initiation of climate models on the other, which improves the accuracy of future predictions. Noise in GRACE Level-2 data (monthly gravity field solutions) is caused by various reasons. The measurements themselves are already executed and their quality is fixed but better data processing algorithms and background models can reduce the current noise level. This is also relevant for the GRACE Follow On mission which might have a higher measurement precision. Over the ocean, these GRACE monthly solutions ideally only show mass exchange between continents and ocean and effects of self-attraction and loading. Therefore, the signal over the ocean is expected to consist predominantly of a linear trend and a seasonal variability. For certain oceanic regions this is not the case. In these areas still a signal variance representing interannual differences in the mass-derivative and large residuals with respect to a low-pass filtered signal are observed. This low-pass filtered signal contains only signals of a frequency lower than the semiannual cycle. These signal variance and residuals are unexpected and can be caused by inaccuracies in the currently applied oceanic background models in GRACE data processing. For various Release 5 and Release 6 monthly solutions the noise variance, signal variance and residuals as aforementioned are estimated. The noise and signal variance are estimated by Variance Component Estimation (VCE). Additionally, numerical experiments are performed to analyze different regularization functionals and set-ups in the VCE. The oceanic regions where the largest signal variance and residuals are observed correlate. These areas are for GRACE Release 5 data the Baltic Sea, Black Sea, Arafura Sea, East Siberian Arctic Shelf, Argentine Basin and Hudson Bay. For GRACE Release 6 data a significant drop of this signal variance and residuals can be observed for the Hudson Bay and East Siberian Arctic Shelf. Consequently, the oceanic background models for these releases are compared against each other and against the Global Tide and Surge Model (GTSM) which is a 2D hydrological model based on the Delft3D Flexible Mesh software developed by Deltares. For the whole ocean both 3/6-hourly time-series and monthly time-series are analyzed. For the shallow regions up to 200 m, the Black Sea and the Red Sea, GTSM shows significant differences with respect to the current applied oceanic background models. When comparing the oceanic background models of different releases it can be observed that the regions where the signal variance and residuals decreased for GRACE Release 6 with respect to Release 5 correlate to regions where the differences between these models is significant. This indicates that oceanic background models do significantly influence the quality of GRACE monthly solutions over the ocean. Furthermore, it is investigated whether it can be expected that GTSM will improve the GRACE monthly solutions. For this, monthly time-series of the current applied oceanic background models are added back to the GRACE monthly solutions; consequently, by GTSM computed monthly time-series are removed. Compared to GRACE Release 5 monthly solutions, GTSM shows a reduction in the signal variance and residuals for the Hudson Bay, East Siberian Arctic Shelf, Black Sea, Baltic Sea, North Sea, Arafura Sea and certain parts of the Arctic and Southern ocean. Compared to GRACE Release 6, a reduction in the signal variance and residuals is observed for the East Siberian Arctic Shelf, Black Sea, Baltic Sea, North Sea and Arafura Sea. For these regions it is most expected that GTSM can improve GRACE monthly solutions. Since the quality of monthly solutions over the oceans is clearly influenced by the oceanic background models significant alterations in GRACE monthly solutions are expected for the shallow regions up to 200 m, Black Sea and Red Sea when applying GTSM in the GRACE data processing. Whether these will be improvements or not should be analyzed by implementing GTSM-based 3-hourly time-series in the GRACE data processing to create a new GRACE Level-2 data product.

Ocean currents play a crucial role in many scientific and industrial applications. Contemporary measurement techniques are limited in spatial coverage or spatial resolution. This study presents a proof-of-concept for a new measurement principle that merges optical satellite imagery of Kelvin wakes with data from the Automatic Identification System (AIS). A case study in the Strait of Gibraltar was performed using two months of Sentinel-2 imagery, which yielded 81 visible Kelvin wakes over 25 images. For each Kelvin wake, currents were estimated in directions parallel and perpendicular to the ship's sailing line. The estimated currents were validated with respect to surface currents derived from High-Frequency Radars (HFRs) and modelled currents from the Copernicus Marine Environmental Monitoring Service (CMEMS). The results suggest that the estimated currents were highly accurate in the absence of large variations in ship course. However, the frequency of measurements is limited by satellite repeat times and Kelvin wake visibility. More research is needed to explore the potential spatiotemporal distribution of measurements.

Around the world, there are 58 Small Island Developing States (SIDS). What these SIDS generally have in common is that they are very susceptible to different natural hazards, among which coastal flooding. In order to address this issue, adaptation measures are needed. In order to prioritize adaptation efforts when considering these islands at a large scale, it is important to obtain an overview of which islands are most vulnerable to coastal flooding. This requires a large-scale assessment of coastal flooding for these islands. Such a large-scale assessment introduces different challenges, that were looked into in the hydraulic engineering part of this thesis. A first challenge is the availability op topography data for these islands. For most islands, only satellite-based DEMs are available. The accuracy of different satellite-based DEMs was assessed by comparing them to more accurate elevation data for 11 different islands. It could be concluded that in general, the TanDEM-X DEM is the most accurate for terrains with milder slopes, and the ALOS DEM for terrains with steeper slopes. Furthermore, the DEM error was found to be strongly correlated with forested and builded areas. Especially the DEM error in builded areas poses a problem for coastal flood assessments, as these are the main areas of interest in such assessments. Therefore, a building-correction method based on data from Open Street Map (OSM) was proposed and implemented. It could be concluded that the building-correction can both in- and decrease the DEM error, depending on the other error sources that are present in the DEM. A second challenge that is introduced when upscaling coastal flood assessments, is the used method for flood modelling. In the context of this thesis, a simple flood model (called the JBIW model) was developed. This model is a combination of the Janssen-Battjes model for short wave dissipation and the IW-method, which was originally developed for large-scale river flooding calculations. The JBIW model was tested in 1D and compared to other flood models. The results indicated that the simple flood model can be used to obtain a first estimate of coastal flooding, but is not accurate enough to predict exact values of the maximum water levels. Furthermore, the JBIW model was implemented in 2D for the island of Ebeye. The model was combined both with accurate elevation data and with building-corrected satellite-based DEMs. The results indicated that the error introduced by the use of the simple JBIW model were much smaller than the errors introduced by the use of the satellite-based DEMs. This indicates that further research focusing on the upscaling of coastal flood assessments for reef-fronted SIDS should focus on obtaining more accurate elevation data for these islands. Apart from allocating resources to the SIDS in the most efficient way, it is important that the resources that are allocated to a certain island are used effectively. There are different ways to help these islands. An interesting approach is to focus on increasing the adaptive capacity of the SIDS. In the context of the science communication part of this thesis it was investigated whether the simple JBIW model could be used to increase the adaptive capacity to coastal flooding in the context of São Tomé. In order to do this, a theoretical framework was developed that aims to provide practical guidance in the assessment of the (barriers to the) adaptive capacity of a certain system. This framework was applied to the context of São Tomé to map the adaptive capacity to coastal flooding of the system and obtain an overview of the most important barriers to this adaptive capacity. Subsequently, it was assessed which barriers could be addressed with a tool based on the JBIW model. These barriers were used as the starting point for an initial design of the tool interface.

Accurate information on Arctic sea ice thickness has been historically limited both spatially and temporally to spare submarine sonar measurements until the advent of satellite altimeters such as ICEsat (operating from 2002 to 2008) and CryoSat-2 (operating from 2011 to present day). This study aims to use the historical record of normalized radar backscatter measurements from satellite Ku-band and C-band scatterometers (ERS, QuikSCAT and ASCAT), which have been continuously operating since 1992, to homogenize the satellite altimeter record and extend the record of Arctic sea ice thickness measurements backwards in time. This study is structured so as to first derive a set of empirical relationships between normalized backscatter measurements and wintertime sea ice thickness estimates in the Arctic using existing satellite altimeter records as a reference. Two separate scatterometer sea ice thickness models are produced using coincident scatterometer and altimeter observations, one for C-band sea ice thickness estimation using ASCAT and CryoSat-2 collocations, and another for Ku-band sea ice thickness estimation using QuikSCAT and ICESat collocations. Based on the agreement to the altimeter records, the estimation of wintertime sea ice thickness using the C-band and Ku-band scatterometer models is uncertain to within 0.5 m (1-sigma), that is, a precision similar to that of the original altimeter references. The homogenization of the satellite altimeter records cannot be done directly, because the ICESat and CryoSat-2 instruments operate in different periods, but it can be done indirectly by comparing the sea ice thickness estimates obtained from Ku-band (based on ICESat) and C-band (based on CryoSat-2) estimates during the years that the Ku-band and C-band scatterometers operate simultaneously. These overlap years have been used to verify the consistency between the C-band and Ku-band relationships, and to correct for a 0.55 m bias in the CryoSat-2 reference, having considered the earlier ICESat record as absolute standard. After removing this bias from the CryoSat-2 reference, the sea ice thickness estimates from C-band and Ku-band records agree to within 0.15 m (1-sigma). Moreover, the resulting scatterometer sea ice thickness models allow the introduction of new thickness thresholds from a previously existing backscatter-based classification of Arctic sea ice types, providing a thickness threshold of 1.54 m to define first year ice (FYI), and a thickness threshold of 2.25 m to separate second year ice (SYI) from older multiyear ice (old MYI). Ancillary datasets were used to investigate the correlations between backscatter-based sea ice thickness and physical variables, such as snow depth and surface deformation, in order to investigate possible sources of systematic error, which otherwise appear to be bound within 0.30 m (1-sigma). The maps of differences between scatterometer and altimeter sea ice thickness estimates were analysed in terms of collocated sea ice convergence, sea ice shear and snow depth parameters using a multiple regression model. The results show that both the Ku-band and C-band models underestimate ice thickness in areas of high convergence such as the Fram Strait, and overestimate ice thickness in areas with high shear such as the Beaufort Gyre. These correlations may be interpreted as led by increases in backscatter due to surface deformation with (case of convergence) or without (case of shear) associated increases in ice thickness. In addition, the Ku-band model is found sensitive to snow load, with overestimation interpreted as led by an increase in backscatter without associated ice growth, and the C-band model is sensitive to marginal rough ice. An unphysically large dependence on snow depth was found for the C-band estimate, which we conjecture is due to problems with the CryoSat-2 reference. Finally, a reconstruction of Arctic sea ice thickness in the wintertime has been made by combining the Ku-band and C-band sea ice thickness models with the normalized radar backscatter record of ERS, QuikSCAT and ASCAT from 1992 through 2017. These showed a decline average Arctic sea ice thickness of -0.28 m / decade, with a steady decline in second-year ice and high variability in the mean multi-year ice thickness.