Identify and interpret fields and derived products
Ján Kanák presents the operational satellite products for precipitation detection, the procedure for their validation and a case study showing the use of these products in evaluating the long-term accumulated precipitation.
Primary satellite data processed into higher-level products are still used less frequently, especially in the context of processing by NWC SAF software, or directly by SAF products received by the EUMETCast Satellite receiving system. Such products include the hydrology support products of the EUMETSAT H SAF (Hydrological Satellite Application Facility). SHMÚ, as a member of the consortium, has long been involved in the task of validation of products for precipitation detection and hydrological applications of these products. In this article we present the operational satellite products for precipitation detection, the procedure for their validation and a case study presenting the use of these products in evaluating the long-term accumulated precipitation. Accumulated precipitation can be used to monitor periods of droughts with precipitation deficits and surpluses. The ambition of this work is to show future users of satellite data that satellite products of a higher level of processing have the potential for climatological studies. A significant increase in this potential is expected in the near future with the launch of the new generation of MTG (third generation Meteosat) and EPS-SG (second generation European Polar System) satellites.
Hamidreza Mosaffa presents a study that aims 1) to develop the long-term climatological SM2RAIN datasets for the period of 1998–2020 by merging two rainfall SM2RAIN products including SM2RAIN-CCI and SM2RAIN-ASCAT, and 2) to the analysis of drought based on standardized precipitation index over the USA.
Investigation of drought variability requires long term rainfall dataset with high spatial and temporal resolution. The goal of this study are as follow: 1) to develop the long-term climatological SM2RAIN datasets for the period of 1998–2020 at 0.25° spatial and monthly temporal resolution by merging two rainfall SM2RAIN products including SM2RAIN-CCI and SM2RAIN-ASCAT, and 2) to the analysis of drought based on standardized precipitation index over the USA. Results indicated that the most significant decreases in the monthly rainfall trends appear in November. In addition, drought occurred during 2003, 2007, and 2012 over most parts of the USA.
Jean-Christophe Calvet presents a neural network that has been trained using the modelled surface soil moisture (SSM), soil temperature, rainwater interception by leaves, and satellite-derived LAI observations from Copernicus.
In the context of climate warming, the frequency and the intensity of extreme events such as droughts is increasing, and better modelling the response of vegetation to climate is needed. Monitoring the impact of extreme events on terrestrial surfaces involves a number of variables of the soil-plant system such as surface albedo, the soil water content and the vegetation leaf area index (LAI). These variables can be monitored by either using the unprecedented amount of data from the Earth observation satellite fleet, or using land surface models. Another solution consists in combining all available sources of information by assimilating satellite observations into models. In this work, level 1 ASCAT backscatter values (sigma0) are assimilated in the ISBA land surface model of Meteo-France using the LDAS-Monde tool. First, an observation operator is built using machine learning. A neural network (NN) is trained using the modelled surface soil moisture (SSM), soil temperature, rainwater interception by leaves, and satellite-derived LAI observations from Copernicus. The NN is then used for simulating sigma0, making LDAS-Monde capable of assimilating ASCAT sigma0 observations. It is shown that the assimilation of sigma0 alone is able to markedly improve the simulated LAI and soil moisture.
Boris Gratadoux presents an automatic flood forecast tool that combines two kinds of information used as input of a Machine Learning algorithm.
Thales has developed an automatic flood forecast tool combining two kinds of information used as input of a Machine Learning algorithm:
- - Forecasted flows at the outlet of the watershed of interest, obtained with an ensemble data assimilation using a particle filter;
- - Current soil state information from space-based observations;
The prototype allows forecasting time series of flood occurrence associated with a confidence index for a 6-day flood forecast with a 6-hour time-step. It has been tested and validated on a French watershed with good results.
Yves Tramblay presents the results of a study that evaluates satellite rainfall products for hydrological modelling over 12 basins in Morocco using four different hydrological models.
Morocco, as many African countries, has limited observed precipitation data that is a major obstacle for water management, flood monitoring and climate change adaptation planning using hydrological models. The objective of this study is to evaluate satellite rainfall products for hydrological modelling over 12 basins in Morocco using four hydrological models: IHACRES, MISDc, GR4J and CREST. Six satellite products are used in this analysis: the H03, H05, H64 and H67 HSAF products, in addition to SM2RAINASCAT product and GPM IMERG-E. The results showed that the best results to reproduce river runoff are achieved with the SM2RAINASCAT and H64 products, using the CREST and MISDc hydrological models. However, there are strong interplays between the different precipitation products and hydrological model structures in different basins, highlighting the need to carefully select hydrological models according to the intended application. This first evaluation over 12 Moroccan basins suggests that the use of satellite rainfall products for hydrological modelling could a viable alternative to observed rainfall in basins where precipitation is not monitored.
Daniele Casella presents a new algorithm for the Micro-Wave Sounder (MWS) radiometer on board the EPS Second Generation satellites.
The development of precipitation retrieval techniques can now benefit from the availability of unique cloud and precipitation observations by the two space borne radars currently available: the Dualfrequency Precipitation Radar (DPR) on board the NASA/JAXA GPM Core Observatory, and the NASA CloudSat Cloud Profiling Radar (CPR). These two radars have demonstrated their complementarity in the monitoring of precipitation. While DPR has shown a high accuracy in the estimate of medium and intense precipitation regimes, CPR has proven to be very suitable for the retrieval of light rain and snowfall. Within the H SAF a new algorithm for the Micro-Wave Sounder (MWS) radiometer on board the EPS Second Generation satellites (MetOp-SG) has been developed. Different machine learning approaches were tested in order to select the most suitable for optimizing the performance of the algorithm for the detection (a classification problem) and estimate (a regression problem) of rainfall and snowfall.
The details concerning the coincidence datasets creation, the design of the ML modules and the algorithm input selection procedure will be presented, together with results of the algorithm's performance.
Giulia Panegrossi presents the challenges and recent advancements in satellite-based snowfall quantification and global monitoring.
Snow plays an important role in the Earth energy exchange processes, and is a fundamental element of the water cycle. The use of satellites for snowfall monitoring and quantification and for retrieving snow cover properties and variability is necessary to globally quantify water resources. Recent studies have evidenced how space borne multi-channel microwave (MW) radiometer measurements respond to both snowfall and snow cover properties. Improvement in both monitoring of high latitude precipitation and in our understanding on microphysical and dynamical processes that influence high latitude precipitation patterns, intensity and type must be driven by concerted observations of active radars and passive microwave radiometers. This has been recently demonstrated through the development of machine learning-based algorithms for snowfall detection and retrieval, exploiting global observational datasets built from passive and active microwave space borne sensors. In this presentation the challenges and recent advancements in satellite-based snowfall quantification and global monitoring will be discussed. Moreover, retrieval strategies based on machine learning approaches that are being adopted within the EUMETSAT H SAF in view of the future EPS-SG mission, will be presented.
Semih Kuter talks about improvement of the EUMETSAT H-SAF H35 effective snow-covered area product by Multivariate Adaptive Regression Splines.
The pre-operational daily H35 effective snow-covered area product of H-SAF spatially spans the Northern Hemisphere and it is the successor of the operational Pan-European H12 product. Both products are at ~1 km resolution and developed through the H-SAF project of EUMETSAT by exploiting AVHRR channels. During the AS activity (i.e., H_AVS_18_03), an alternative machine learning-based approach is applied on H35 product to improve its accuracy. The new version of H35 product is realized through multivariate adaptive regression splines (MARS) algorithm. AVHRR reflectance data, as well as the well-known snow and vegetation indices (i.e., NDSI and NDVI), are used as predictors to generate the new MARS-based H35 product. The reference fractional snow-covered area (fSCA) maps are obtained from higher resolution Sentinel 2 imagery. Rigorous assessment on the final MARS-based H35 is performed over the Northern Hemisphere within a temporal domain from Nov 2018 to Nov 2019 by using i) Sentinel 2 derived reference fSCA maps, ii) ERA5-Land snow depth data, iii) MODIS MOD10A1 NDSI snow cover data, and iv) in-situ snow depth data. An additional visual assessment is also carried out by comparing MARS-H35/MODIS false-color and MARS-H35/Sentinel 2-derived reference fSCA image pairs over various geographic regions.
Filipe Aires presents statistical approaches to assimilate soil moisture information.
Land surfaces are characterized by strong heterogeneities of soil texture, orography, land cover, soil moisture, snow and other variables. These are very challenging to represent accurately in radiative transfer models, which currently still have limited reliability over land. In this study, we compare two statistical modeling approaches: the traditional CDF-matching used routinely in NWP centers (used here as a normalization and as an inversion technique), and the Neural Network (NN) methods. NNs and CDF-matching are compared and combined. Two cases are considered: (1) the more traditional inversion scheme, and (2) the forward modelling that could be attractive for assimilation purposes. It is shown that in the context of ASCAT, the inversion approach seems better suited than the forward modelling but this could be different for another type of observations. It is also shown that it is possible to combine the global model obtained using the NN and the localized information of the LSM offered by the CDF-matching. A first assessment is performed over the US using in situ soil measurements. Finally, we will present future plans to develop a forward operator for low-frequency microwave channels (SMOS, AMSR-E, SMAP, CIMR) based on a statistical modeling of surface emissivities over continental, snow-ice and sea ice surfaces.
Annalina Lombardi demonstrates that that areal precipitation estimation including satellite information in addition to surface observations has a high performance compared to that which uses only the rain gauge data.
Satellite-based remote sensing provides a significant contribution for hydrological predictions due to its wide coverage and increasing tempo-spatial resolutions. Although current observed areal precipitation estimation is mainly based on point rain gauge measurement interpolation, the ability to deduce spatially distributed data from point measurement depends on the design and density of the sensor network.
A possible approach to have a correct representation of the rain field at the hydrological scale (up to a few hundred meters) could be to merge rain gauge data with gridded rainfall data obtained from remote sensing techniques, and the availability of such data in near-real time is a unique opportunity for the operational hydrology community. In this study we propose the application of advanced downscaling techniques based on Cellular Automata Algorithm for rainfall spatialisation using satellite precipitation products for hydrological applications. The method proposed to merge rainfall estimates measured in different spatial scales is based on the data assimilation concepts with particular emphasis on the transformation of point data to areal data. The work wants to prove that the areal precipitation estimation including satellite information in addition to surface observations has a high performance compared to that which uses only the rain gauge data.
Leo Pio D'Adderio analyses the performances of the H SAF precipitation products during the Mediterranean cyclone Apollo.
This work analyses the performances of the H SAF precipitation products during the Mediterranean cyclone Apollo. The cyclone Apollo occurred between October 25th and October 29th, 2021 with its maximum on October 28th and 29th when it approached the coasts of Sicily causing floods and damages to civil structures. The present work aims to describe the cyclone evolution by exploiting the satellite-based Level 2 and Level 3 H SAF precipitation products. The analysis focuses on the reconstruction of the precipitation pattern and of the quantitative amount thanks a direct comparison with ground-based measurements.
Emanuel Dutra presents an overview of the LSA SAF ET products with particular focus on the use of HSAF soil moisture in operations (H14 and future H26) and in the production of the climate data record (H141).
Evapotranspiration (ET) is a key component of the Earth System coupling land-atmosphere exchanges of water, energy and carbon. Due to the complex nature of the processes involved (radiation, boundary layer turbulence, vegetation status, soil moisture) it is very challenging to observe (locally and from Satellite) and to model. LSA SAF evapotranspiration product takes advantage of satellite radiation and soil moisture estimates to diagnose ET using a surface energy balance approach. This talk will present an overview of the LSA SAF ET products with particular focus on the use of HSAF soil moisture in operations (H14 and future H26) and in the production of the climate data record (H141).