Optimizing a backscatter forward operator using Sentinel-1 data over irrigated land

Worldwide, the amount of water used for agricultural purposes is rising, and the quantification of irrigation is becoming a crucial topic. Because of the limited availability of in situ observations, an increasing number of studies is focusing on the synergistic use of models and satellite data to detect and quantify irrigation. The parameterization of irrigation in large-scale land surface models (LSMs) is improving, but it is still hampered by the lack of information about dynamic crop rotations, or the extent of irrigated areas, and the mostly unknown timing and amount of irrigation. On the other hand, remote sensing observations offer an opportunity to fill this gap as they are directly affected by, and hence potentially able to detect, irrigation. Therefore, combining LSMs and satellite information through data assimilation can offer the optimal way to quantify the water used for irrigation. This work represents the first and necessary step towards building a reliable LSM data assimilation system which, in future analysis, will investigate the potential of high-resolution radar backscatter observations from Sentinel-1 to improve irrigation quantification. Specifically, the aim of this study is to couple the Noah-MP LSM running within the NASA Land Information System (LIS), with a backscatter observation operator for simulating unbiased backscatter predictions over irrigated lands. In this context, we first tested how well modelled surface soil moisture (SSM) and vegetation estimates, with or without irrigation simulation, are able to capture the signal of aggregated 1 km Sentinel-1 backscatter observations over the Po Valley, an important agricultural area in northern Italy. Next, Sentinel-1 backscatter observations, together with simulated SSM and leaf area index (LAI), were used to optimize a Water Cloud Model (WCM), which will represent the observation operator in future data assimilation experiments. The WCM was calibrated with and without an irrigation scheme in Noah-MP and considering two different cost functions. Results demonstrate that using an irrigation scheme provides a better calibration of the WCM, even if the simulated irrigation estimates are inaccurate. The Bayesian optimization is shown to result in the best unbiased calibrated system, with minimal chances of having error cross-correlations between the model and observations. Our time series analysis further confirms that Sentinel-1 is able to track the impact of human activities on the water cycle, highlighting its potential to improve irrigation, soil moisture, and vegetation estimates via future data assimilation.

Towards All-sky Assimilation of Microwave Temperature Sounding Channels in Environment Canada’s Global Deterministic Weather Prediction System

The all-sky assimilation of radiances from microwave instruments is developed in the 4D-EnVar analysis system at Environment and Climate Change Canada (ECCC). Assimilation of cloud-affected radiances from Advanced Microwave Sounding Unit A (AMSUA) temperature sounding channels 4 and 5 for non-precipitating scenes over the ocean surface is the focus of this study. Cloud-affected radiances are discarded in the ECCC operational data assimilation system due to the limitations of forecast model physics, radiative transfer models, and the strong non-linearity of the observation operator. In addition to using symmetric estimate of innovation standard deviation for quality control, a state-dependent observation error inflation is employed at the analysis stage. The background state clouds are scaled by a factor of 0.5 to compensate for a systematic overestimation by the forecast model, before being used in the observation operator. The changes in the fit of the background state to observations show mixed results. The number of AMSUA channels 4 and 5 assimilated observations in the all-sky experiment is 5-12% higher than in the operational system. The all-sky approach improves temperature analysis when verified against ECMWF operational analysis in the areas where the extra cloud-affected observations were assimilated. Statistically significant reductions in error standard deviation by 1-4% for the analysis and forecasts of temperature, specific humidity, and horizontal wind speed up to maximum 4 days were achieved in the all-sky experiment in the lower troposphere. These improvements result mainly from the use of cloud information for computing the observation-minus-background departures. The operational implementation of all-sky assimilation is planned for Fall 2021.

Multi-sensor analyses of the skin temperature for the assimilation of satellite radiances in the European Centre for Medium-Range Weather Forecasts (ECMWF) Integrated Forecasting System (IFS, cycle 47R1)

The assimilation of clear-sky radiance in the European Centre for Medium-Range Weather Forecasts (ECMWF) atmospheric analysis relies on the clear-sky radiances observation operator. Some of these radiances have frequencies that make them sensitive to both the surface and atmosphere. Because the atmospheric and surface analyses are currently not strongly coupled, a specific treatment of the surface is required. The observation operator specifically expects a skin temperature value at the observation location and time as well as the profiles of the atmospheric variables along the viewing path. This skin temperature is added to the control variable and optimised simultaneously with all of the atmospheric variables to produce optimal simulated radiances. We present two approaches to add the skin temperature to the control variable. In the current TOVS Control Variable (TOVSCV) approach, a series of skin temperature values per observation location is added to the control variable. Effectively, in the optimisation process, the skin temperature acts as a sink variable in observation space and is uncoupled from the skin temperature at other locations. In the novel SKin Temperature in the Extended Control Vector (SKTECV) approach, two-dimensional skin temperature fields are added to the control variable. All clear-sky radiances then participate in the optimisation of these two-dimensional fields, and the analysis produces temporally and spatially consistent skin temperature fields. We compare the two approaches over two seasons of 3 months each. Overall, there is a neutral impact of the new approach on the analysis and forecast. Moreover, there is some evidence that the contribution of the subsurface layers should be represented in the new approach for the skin temperature associated with the microwave instruments.

Optimizing a backscatter forward operator using Sentinel-1 data over irrigated land

Worldwide, the amount of water used for agricultural purposes is rising and the quantification of irrigation is becoming a crucial topic. Because of the the limited availability of in situ observations, an increasing number of studies is focusing on the synergistic use of models and satellite data to detect and quantify irrigation. The parameterization of irrigation in large scale Land Surface Models (LSM) is improving, but it is still hampered by the lack of information about dynamic crop rotations or the extent of irrigated areas, and the mostly unknown timing and amount of irrigation. On the other hand, remote sensing observations offer an opportunity to fill this gap as they are directly affected by, and hence potentially able to detect, irrigation. Therefore, combining LSMs and satellite information through data assimilation can offer the optimal way to quantify the water used for irrigation. The aim of this study is to optimize a land modeling system, consisting of the Noah-MP LSM, coupled with a backscatter observation operator, over irrigated land in order to simulate backscatter predictions. This is a first step towards building a reliable data assimilation system to ingest level-1 Sentinel-1 observations. In this context, we tested how well modeled soil moisture and vegetation estimates from the Noah-MP LSM running within the NASA Land Information System (LIS), with or without irrigation simulation, are able to capture the signal of high-resolution Sentinel-1 backscatter observations over the Po river Valley, an important agricultural area in Northern Italy. Next, aggregated 1-km Sentinel-1 backscatter observations were used to calibrate a Water Cloud Model (WCM) as observation operator using simulated soil moisture and Leaf Area Index estimates. The WCM was calibrated with and without activating an irrigation scheme in Noah-MP and considering two different cost functions. Results demonstrate that activating an irrigation scheme provides the optimal calibration of the WCM, even if the irrigation estimates are inaccurate. The Bayesian optimization is shown to result in the best unbiased calibrated system, with minimal chance of having error cross correlations between the model and observations. Our time series analysis further confirms that Sentinel-1 is able to track the impact of human activities on the water cycle, highlighting its potential to improve irrigation, soil moisture and vegetation estimates via future data assimilation.

A Neural Network-Based Observation Operator for Coupled Ocean-Acoustic Variational Data Assimilation

Variational data assimilation requires implementing the tangent-linear and adjoint (TA/AD) version of any operator. This intrinsically hampers the use of complicated observations. Here, we assess a new data-driven approach to assimilate acoustic underwater propagation measurements (Transmission Loss, TL) into a regional ocean forecasting system. TL measurements depend on the underlying sound speed fields, mostly temperature, and their inversion would require heavy coding of the TA/AD of an acoustic underwater propagation model. In this study, the non-linear version of the acoustic model is applied to an ensemble of perturbed oceanic conditions. TL outputs are used to formulate both a statistical linear operator based on canonical correlation analysis (CCA), and a neural network-based (NN) operator. For the latter, two linearization strategies are compared, the best-performing one relying on reverse-mode automatic differentiation. The new observation operator is applied in data assimilation experiments over the Ligurian Sea (Mediterranean Sea), using the Observing System Simulation Experiments (OSSE) methodology to assess the impact of TL observations onto oceanic fields. TL observations are extracted from a nature run with perturbed surface boundary conditions and stochastic ocean physics. Sensitivity analyses indicate that the NN reconstruction of TL is significantly better than CCA. Both CCA and NN are able to improve the upper ocean skill scores in forecast experiments, with NN outperforming CCA on the average. The use of the NN observation operator is computationally affordable, and its general formulation appears promising for the adjoint-free assimilation of any remote sensing observing network.

Multi-sensor analyses of the skin temperature for the assimilation of satellite radiances in the European Centre for Medium-Range Weather Forecasts (ECMWF) Integrated Forecasting System (IFS, cycle 47R1)

The assimilation of clear-sky radiance in the European Centre for Medium-Range Weather Forecasts (ECMWF) atmospheric analysis relies on the clear-sky radiances observation operator. Some of these radiances have frequencies that make them sensitive to both the surface and atmosphere. Because the atmospheric and surface analyses are currently not strongly coupled, a specific treatment of the surface is required. The observation operator expects in particular, a skin temperature value at the observation location and time, together with the profiles of the atmospheric variables along the viewing path. This skin temperature is added to the control variable and optimised simultaneously with all the atmospheric variables to produce optimal simulated radiances.We present two approaches to add the skin temperature to the control variable. In the current TOVSCV approach, a series of skin temperature value per observation location is added to the control variable. Effectively, in the optimisation process, the skin temperature acts as a sink variable in observation space and is uncoupled from the skin temperature at other locations. In the novel SKTACV approach, two-dimensional skin temperature fields are added to the control variable. All clear-sky radiances then participate in the optimisation of these two-dimensional fields and the analysis produces temporally and spatially consistent skin temperature fields.We compare the two approaches over two seasons of three months each. Overall, there is a neutral impact of the new approach on the analysis and forecast. Besides, there are some evidences that the contribution of the sub-surface layers should be represented in the new approach for the skin temperature associated with the microwave instruments.

A Feature Track Correction (FTC) Observation Operator applied to Aeolus-AMV Collocations

<p>In this study we propose and test a feature track correction (FTC) observation operator for atmospheric motion vectors (AMVs).  The FTC has four degrees of freedom corresponding to wind speed multiplicative and additive corrections (γ and δ<em><strong>V</strong></em>), a vertical height assignment correction (<em>h</em>), and an estimate of the depth of the layer that contributes to the AMV (Δ<em>z</em>).  Since the effect of the FTC observation operator is to add a bias correction to a weighted average of the profile of background winds an alternate formulation is in terms of a profile of weights (<em>w<sub>k</sub></em>) and δ<em><strong>V </strong></em>.</p><p>The FTC observation operator is tested in the context of a collocation study between AMVs projected onto the collocated Aeolus horizontal line-of-sight (HLOS) and the Aeolus HLOS wind profiles.  This is a prototype for an implementation in a variational data assimilation system and here the Aeolus profiles act as the background in the FTC observation operator.  Results were obtained for ten days of data using modest QC.  The overall OMB or collocation difference SD for a global solution applied to the independent sample is 5.49 m/s with negligible mean.  For comparison the corresponding simple (or pure) collocation SD is 7.85 m/s, and the null solution, which only interpolates the Aeolus profile to the reported height of the AMV and removes the overall bias, has an OMB SD of 7.23 m/s. These values correspond to reductions of variance of 51.0% and 42.3%, due to the FTC observation operator in comparison to the simple collocation and null solution, respectively.</p><p>These preliminary tests demonstrate the potential for the FTC observation operator for </p><ul><li>Improving AMV collocations (including triple collocation) with profile wind data.</li> <li>Characterizing AMVs. For example, summary results for the HLOS winds show that AMVs compare best with wind profiles averaged over a 4.5 km layer centered 0.5 km above the reported AMV height.</li> <li>Improving AMV observation usage within data assimilation (DA) systems. Lower estimated error and more realistic representation of AMVs with variational FTC (VarFTC) should result in greater information extracted.  The FTC observation operator accomplishes this by accounting for the effects of <em>h</em> and Δ<em>z</em>. </li> </ul>

A machine learning approach to the observation operator for satellite radiance data assimilation

<p>The ‘observation operator’ is essential in data assimilation (DA) to derive the model equivalent of the observations from the model variables. For satellite radiance observations, it is usually based on complex radiative transfer model (RTM) with a bias correction procedure. Therefore, it usually takes time to start using new satellite data after launching the satellites. Here we take advantage of the recent fast development of machine learning (ML) which is good at finding the complex relationships within data. ML can potentially be used as the ‘observation operator’ to reveal the relationships between the model variables and the observations without knowing their physical relationships. In this study, we test with the numerical weather prediction system composed of the Nonhydrostatic Icosahedral Atmospheric Model (NICAM) and the Local Ensemble Transform Kalman Filter (LETKF). We focus on the satellite microwave brightness temperature (BT) from the Advanced Microwave Sounding Unit-A (AMSU-A). Conventional observations and AMSU-A data were assimilated every 6 hours. The reference DA system employed the observation operator based on the RTTOV and an online bias correction method.</p><p>We used this reference system to generate 1-month data to train the machine learning model. Since the reference system includes running a physically-based RTM, we implicitly used the information from RTM for training the ML model in this study, although in our future research we will explore methods without the use of RTM. The machine learning model is artificial neural networks with 5 fully connected layers. The input of the ML model includes the NICAM model variables and predictors for bias correction, and the output of the ML model is the corresponding satellite BT in 3 channels from 5 satellites. Next, we ran the DA cycle for the same month the following year to test the performance of the ML model. Two experiments were conducted. The control experiment (CTRL) was performed with the reference system. In the test experiment (TEST), the ML model was used as the observation operator and there is no separate bias correction procedure since the training includes biased differences between the model and observation. The results showed no significant bias of the simulated BT by the ML model. Using the ECMWF global atmospheric reanalysis (ERA-interim) as a benchmark to evaluate the analysis accuracy, the global-mean RMSE, bias, and ensemble spread for temperature in TEST are 2% higher, 4% higher, and 1% lower respectively than those in CTRL. The result is encouraging since our ML can emulate the RTM. The limitation of our study is that we rely on the physically-based RTM in the reference DA system, which is used for training the ML model. This is the first result and still preliminary. We are currently considering other methods to train the ML model without using the RTM at all.</p>

Outcomes of the Aerosol Radiance Assimilation Study

<p>The ESA-funded Aerosol Radiance Assimilation Study (ARAS) has provided ground-breaking research in using visible radiance data from satellite to estimate the concentration of aerosols.</p><p>Satellite observations in the infrared and microwave parts of the spectrum have long been assimilated into forecasting systems to help estimate the best possible initial conditions for global weather predictions. Assimilating radiances in the visible part of the spectrum, on the other hand, continues to pose many challenges.The reason lies in the complex interactions of cloud and aerosol particles with radiation at those wavelengths and in the complex characteristics of the surface as a reflector of visible light. These factors make it difficult to develop an observation operator, which converts model values into satellite observation equivalents.</p><p>One of the key achievements of ARAS is to have developed an observation operator for aerosol reflectances in the visible part of the spectrum. This operator was comprised of two elements: a fast radiative transfer code based on a Look-Up-Table approach developed by RAL Space for aerosol retrievals (Thomas et al, 2009) and adapted to the ECMWF’s Integrated Forecast System as well as a surface reflectance model for ocean and land.</p><p>This enabled the first-ever experimental assimilation of reflectances into the 4D-Var assimilation system of ECMWF’s Integrated Forecasting System (IFS) to help estimate aerosol concentrations. The assimilation experiments were very successful. The performance was remarkable considering that this was a new development rolled out over the course of just two years. The observations used in the ARAS project were cloud-cleated aerosol reflectances from the MODIS instrument on board the Aqua and Terra satellites. Experiments were carried out to compare the impact of assimilating these observations with the impact of assimilating traditional satellite-derived AOD observations. The results show that the performance of reflectance assimilation is broadly comparable to that of satellite AOD assimilation. However, it varies depending on the metrics used and the period analysed.</p><p>While assimilating aerosol reflectances is still experimental, the results show great potential for future operational implementation in atmospheric composition forecasts. Such forecasts are routinely produced by the EU‐funded Copernicus Atmosphere Monitoring Service (CAMS) implemented by ECMWF. However, the scope for future applications is much wider than that. Many of the tools developed in ARAS for aerosol visible reflectance assimilation could be adapted to clouds. This could open the way towards a fuller exploitation of visible radiances to improve numerical weather prediction.</p><p><strong>References</strong></p><p>Thomas G.E., Carboni E., Sayer A.M., Poulsen C.A., Siddans R., Grainger R.G. (2009) Oxford-RAL Aerosol and Cloud (ORAC): aerosol retrievals from satellite radiometers. In: Kokhanovsky A.A., de Leeuw G. (eds) Satellite Aerosol Remote Sensing over Land. Springer Praxis Books. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-540-69397-0_7</p>