Rainfall Estimates Research Articles

Interpretable Quality Control of Sparsely Distributed Environmental Sensor Networks Using Graph Neural Networks

Abstract Environmental sensor networks play a crucial role in monitoring key parameters essential for understanding Earth’s systems. To ensure the reliability and accuracy of collected data, effective quality control (QC) measures are essential. Conventional QC methods struggle to handle the complexity of environmental data. Conversely, advanced techniques such as neural networks, are typically not designed to process data from sensor networks with irregular spatial distribution. In this study, we focus on anomaly detection in environmental sensor networks using graph neural networks, which can represent sensor network structures as graphs. We investigate its performance on two datasets with distinct dynamics and resolution: commercial microwave link (CML) signal levels used for rainfall estimation and SoilNet soil moisture measurements. To evaluate the benefits of incorporating neighboring sensor information for anomaly detection, we compare two models: graph convolution network (GCN) and a graph-less baseline: long short-term memory (LSTM). Our robust evaluation through 5-fold cross-validation demonstrates the superiority of the GCN models. For CML, the mean area under receiver operating characteristic curve for the GCN was 0.941 compared to 0.885 for the baseline-LSTM, and for SoilNet, it was 0.858 for GCN and 0.816 for the baseline-LSTM. Visual inspection of CML time series revealed that the GCN proficiently classified anomalies and remained resilient against rain-induced events often misidentified by the baseline-LSTM. However, for SoilNet, the advantage of GCN was less pronounced, likely due to inconsistent and less precise labeling. Through interpretable model analysis, we demonstrate how feature attributions vividly illustrate the significance of neighboring sensor data, particularly in distinguishing between anomalies and expected changes in signal level in the time series.

Comparative Trend Analysis of Precipitation Indices in Several Towns of the Sirba River Catchment (Burkina Faso) from CHIRPS and TAMSAT Rainfall Estimates

The increasingly frequent pluvial flood of West African urban settlements indicates the need to investigate the drivers of local rainfall changes. However, meteorological stations are few, unevenly distributed, and work irregularly. Daily satellite rainfall datasets can be used. Nevertheless, these products often need to be more accurate due to sensor errors and limitations in retrieval algorithms. The problem is, therefore, how to characterize rainfall where there is a need for ground-based rainfall records or incomplete series. This study aims to characterize urban rainfall using two satellite datasets. The analysis was carried out in the Sirba river catchment, Burkina Faso, using the Climate Hazards Group InfraRed Precipitation with Station data (CHIRPS) and the Tropical Applications of Meteorology using SATellite and ground-based data (TAMSAT) datasets. Ten indices from the Expert Team on Climate Change Detection and Indices (ETCCDI) of precipitation were calculated, and their statistical trends were evaluated from 1983 to 2023. The study introduces two key innovations: a comparative analysis of precipitation trends using two satellite datasets and applying this analysis to towns within a previously understudied 39,138 km2 catchment area that is frequently flooded. Both datasets agree on the increase of (i) annual cumulative rainfall over all towns, (ii) five-day maximum rainfall over the town of Manni, (iii) rainfall due to very wet days in Gayéri, (iv) days of heavy rainfall in Bogandé, Manni and Yalgho, and (v) days of very heavy rainfall in Yalgho. These findings suggest the need for targeted pluvial flood prevention measures in towns with increasing trends in heavy rainfall.

Investigate ensemble machine learning models to reduce the daily Mean field bias of radar rainfall estimates derived from ZR relationships in the sub-river basins in the middle of Thailand

Investigate ensemble machine learning models to reduce the daily Mean field bias of radar rainfall estimates derived from ZR relationships in the sub-river basins in the middle of Thailand

Relating Rainfall Retrieval Parameters to Network and Environmental Features to Improve Rainfall Estimates from Commercial Microwave Links in the Tropics

Abstract Potentially, the greatest benefit of commercial microwave links (CMLs) as opportunistic rainfall sensors lies in regions that lack dedicated rainfall sensors, most notably low- and middle-income countries. However, current CML rainfall retrieval algorithms are predominantly tuned and applied to (European) CML networks in temperate or Mediterranean climates. This study investigates whether local quantitative precipitation estimates from CMLs in a tropical region, specifically Sri Lanka, can be improved by optimizing two dominant parameters in the rainfall retrieval algorithm RAINLINK, namely, the wet antenna attenuation correction factor Aa and the relative contribution of minimum and maximum received signal levels α. Using a grid search, based on 10 months of CML data from 22 link–gauge clusters consisting of 105 sublinks that lie within 1 km of a daily rain gauge, the optimal values of Aa and α are first derived for the entire country and compared to the default RAINLINK values. Subsequently, the CMLs are grouped by link length, frequency, climate zone, and daily rainfall depth classes, and Aa and α are derived for each of these classes. Calibrating parameters on all clusters across the country only leads to minor improvements. The actual optimal Aa and α values depend on the performance metric favored. Calibrating on network properties, particularly short link length and high-frequency classes, does significantly improve rainfall estimates. By relating the optimal Aa and α values to known network metadata, the results from this study are potentially applicable to other tropical CML networks that lack nearby reference rainfall data. Significance Statement The purpose of this study is to improve rainfall estimates from commercial microwave links in Sri Lanka by optimizing two important rainfall retrieval algorithm parameters. Our results show that relating the optimal parameter values to operating frequency and pathlength improves rainfall estimates more than applying a single optimal parameter set to the entire network. By relating the optimal parameter values to readily known network properties, we aim to make these results applicable to other tropical countries, particularly low- and middle-income countries, that lack adequate reference rainfall data to calibrate rainfall estimates from commercial microwave links on.

Evaluation of Satellite-Based Rainfall Estimates and Application to Monitor Flash Floods in Wadi Araba Region Western Suez Gulf

Evaluation of Satellite-Based Rainfall Estimates and Application to Monitor Flash Floods in Wadi Araba Region Western Suez Gulf

A novel validation of satellite soil moisture using SM2RAIN-derived rainfall estimates

Despite the importance of soil moisture (SM) in various applications and the need to validate satellite SM products, the current in situ SM network is still inadequate, even for developed country such as the United States. Recently, SM2RAIN (Soil Moisture to Rain) algorithm has prominently emerged as a bottom-up approach to derive rainfall data from SM. In this study, we evaluated whether SM2RAIN algorithm and rain gauges, which are more abundant and readily available than in situ SM, can be used to validate satellite-based SMAP SM estimates. Since errors in SMAP SM propagate to SMAP-derived rainfall, the skills of SM2RAIN might be able to provide insights on the accuracy of SMAP SM observations. While the correlation between SM2RAIN skills and SMAP SM skills was found to be statistically significant, the strength of the correlation varied among different climate zones and annual rainfall classes. Specifically, weaker correlations were observed in arid and lower rainfall regions (median R value of 0.12), while stronger correlations were found in temperate and higher rainfall regions (median R value of 0.54). In term of over/under-estimation tendencies, 56% of the stations had the same tendencies (SM2RAIN skills and satellite SM skills both have positive or negative PBIAS value).

Evaluation of WSR-88D Level III and MRMS Rainfall Estimates Against Rain Gauge Observations at the Savannah River Site

Rainfall data for the Savannah River Site (SRS) has been historically measured by rain gauges. These instruments serve as ground truth for most climatological and weather applications; however, gauge measurements are prone to errors or biases under certain weather conditions. Rainfall estimates from radar reflectivity values have been developed and improved over the years and serve as an alternative or supplement for gauge measurements. This study compares measurements from tipping bucket rain gauges with rainfall estimates from the NOAA NWS WSR-88D Level III hourly rainfall product and the Multi-Radar Multi-Sensor (MRMS) gauge-corrected precipitation estimate. Results show good agreement between radar derived amounts and ground measurements, with MRMS values showing correlation coefficients of 0.60-0.95 and RMSE values of less than 1.0 cm (0.4 in). The WSR-88D Level III estimates result in correlation coefficients of 0.46-0.86 and RMSE values of less than 1.3 cm (0.5 in). Few outliers are observed for each data pair and are evaluated against precipitation classification products (WSR-88D Hybrid Hydrometeor Classification product and MRMS Precipitation Flag product). The rain gauges used in this study are not part of the Hydrometeorological Automated Data System (HADS) network used to correct the MRMS estimates and therefore, results of this work provide an independent validation of the MRMS gauge correction scheme.

Extreme rainfall in Dakar (Senegal): a case study for September 5, 2020

West African countries frequently experience extreme rainfall events during the monsoon season. On September 5, 2020, a significant event occurred in the Dakar region of Senegal with daily rainfall totals exceeding 90 mm, causing widespread flooding and displacing 1,000's of people. Despite the severity of this event, the physical mechanisms driving such extreme rainfall remain unexplored. This study aims to investigate the physical mechanisms associated with this event using multiple data sources, including satellite rainfall estimate products (GPM-IMERG, CHIRPS) and reanalysis data (ERA-5). By analyzing wind fields and mid-tropospheric moisture content from reanalysis data, we examined the synoptic-dynamic evolution of the atmosphere and the movement of the cyclonic vortex that transported moisture to the affected region, resulting in substantial rainfall measurements exceeding 100 mm. The analysis also revealed that a vortex over the ocean slowed down the vortex near Senegal, prolonging the rainfall over a total period of 10 h. Additionally, this study presents a comprehensive comparative analysis of state-of-the-art satellite rainfall estimates, assessing their accuracy and reliability in capturing extreme rainfall events both spatially and at specific rainfall gauges situated in Dakar. This evaluation revealed that while satellite rainfall estimates are valuable, they tended to underestimate (up to 40%) the actual rainfall observed at the Dakar-Yoff station. Furthermore, extreme value analysis showed that there is a tendency to underestimate return levels for high-intensity events, with some cases showing underestimations by up to twice the actual values. Thus, this research advances our understanding of extreme rainfall events in West Africa and improves our knowledge of satellite-based rainfall estimates, contributing to future monitoring and preparedness. Furthermore, these findings highlight the importance of monitoring cyclonic systems associated with African Easterly Waves, contributing to a better understanding of extreme rainfall events in West Africa.

Uncertainty analysis for design rainfall estimation using peaks-over-threshold model and specially formulated pivotal quantities

Uncertainty analysis for design rainfall estimation using peaks-over-threshold model and specially formulated pivotal quantities

A Localized Quantitative Precipitation Estimation for S-Band Polarimetric Radar in Taiwan

Abstract A polarimetric radar quantitative precipitation estimation (QPE) to estimate the rain rate R from specific attenuation A has been applied in Taiwan’s operational Quantitative Precipitation Estimation and Segregation Using Multiple Sensors (QPESUMS) system since 2016. A 3-yr (2016–18) drop size distribution (DSD) dataset from an operational Particle Size and Velocity (Parsivel) network was used to derive a localized coefficient as well as the α(K) function in the R(A) scheme for S-band radar, where α is a key parameter in the estimation of A and K is the linear fitted slope of differential reflectivity ZDR versus reflectivity Z. The local drop size distribution data were also used to derive the localized R(Z) and R(KDP) relationships, and the relationships were evaluated using radar observations in heavy rain cases. A synthetic quantitative precipitation estimate combining the localized R(A), R(Z), and R(KDP) relationships is compared to its operational counterpart and showed about 8% reduction in the normalized mean error for the mei-yu cases. Typhoon cases exhibited similar improvements by the localized QPE relationships but showed higher uncertainties than in the mei-yu cases. The higher uncertainties in the typhoon QPE verification were likely due to the stronger winds in typhoons than in the mei-yu events that caused greater mismatches between the radar observations at an altitude and the gauges at the ground. Overall, the results demonstrated advantages of localized radar rainfall relationships derived from the disdrometer data to improve the accuracy of the operational rainfall estimation products. Significance Statement A 3-yr (2016–18) drop size distribution (DSD) dataset from the operational Parsivel network in Taiwan was utilized to localize parameters in the QPE relationships for S-band dual-polarimetric radar. These relationships were evaluated using radar observations in the mei-yu front and typhoon events in Taiwan. The results show that using localized radar rainfall relationships derived from the disdrometer data can enhance the accuracy of the operational rainfall products.

Deep Learning for Opportunistic Rain Estimation via Satellite Microwave Links.

Accurate precipitation measurement is critical for managing flood and drought risks. Traditional meteorological tools, such as rain gauges and remote sensors, have limitations in resolution, coverage, and cost-effectiveness. Recently, the opportunistic use of microwave communication signals has been explored to improve precipitation estimation. While there is growing interest in using satellite-to-earth microwave links (SMLs) for machine learning-based precipitation estimation, direct rainfall estimation from raw signal-to-noise ratio (SNR) data via deep learning remains underexplored. This study investigates a range of machine learning (ML) approaches, including deep learning (DL) models and traditional methods like gradient boosting machine (GBM), for estimating rainfall rates from SNR data collected by interactive satellite receivers. We develop real-time models for rainfall detection and estimation using downlink SNR signals from satellites to user terminals. By leveraging a year-long dataset from multiple locations-including SNR measurements paired with disdrometer and rain-gauge data-we explore and evaluate various ML models. Our final models include ensemble approaches for both rainfall detection and cumulative rainfall estimation. The proposed models provide a reliable solution for estimating precipitation using Earth-satellite microwave links, potentially improving precipitation monitoring. Compared to the state-of-the-art power-law-based models applied to similar datasets reported in the literature, our ML models achieve a 46% reduction in the root mean squared error (RMSE) for event-based cumulative precipitation predictions.

Comparison of Different Quantitative Precipitation Estimation Methods Based on a Severe Rainfall Event in Tuscany, Italy, November 2023

Accurate quantitative precipitation estimation (QPE) is fundamental for a large number of hydrometeorological applications, especially when addressing extreme rainfall phenomena. This paper presents a comprehensive comparison of various rainfall estimation methods, specifically those relying on weather radar data, rain gauge data, and their fusion. The study evaluates the accuracy and reliability of each method in estimating rainfall for a severe event that occurred in Tuscany, Italy. The results obtained confirm that merging radar and rain gauge data outperforms both individual approaches by reducing errors and improving the overall reliability of precipitation estimates. This study highlights the importance of data fusion in enhancing the accuracy of QPE and also supports its application in operational contexts, providing further evidence for the greater reliability of merging methods.

Assessing rainfall radar errors with an inverse stochastic modelling framework

Abstract. Weather radar is a crucial tool for rainfall observation and forecasting, providing high-resolution estimates in both space and time. Despite this, radar rainfall estimates are subject to many error sources – including attenuation, ground clutter, beam blockage and drop-size distribution – with the true rainfall field unknown. A flexible stochastic model for simulating errors relating to the radar rainfall estimation process is implemented, inverting standard weather radar processing methods and imposing path-integrated attenuation effects, a stochastic drop-size-distribution field, and sampling and random errors. This can provide realistic weather radar images, of which we know the true rainfall field and the corrected “best-guess” rainfall field which would be obtained if they were observed in a real-world case. The structure of these errors is then investigated, with a focus on the frequency and behaviour of “rainfall shadows”. Half of the simulated weather radar images have at least 3 % of their significant rainfall rates shadowed, and 25 % have at least 45 km2 containing rainfall shadows, resulting in underestimation of the potential impacts of flooding. A model framework for investigating the behaviour of errors relating to the radar rainfall estimation process is demonstrated, with the flexible and efficient tool performing well in generating realistic weather radar images visually for a large range of event types.

Spatial evaluation rainfall estimation on weather radar using Marshall–Palmer reflectivity–rainfall rate in Banten

Based on data from the National Disaster Management Agency (Ind.: Badan Nasional Penanggulangan Bencana – BNPB), throughout 2022, more than 91% of disaster events were hydrometeorological disasters, with floods at 43% and landslides at 17%. One of the factors for floods and landslides is high rainfall intensity. Automatic rain gauge (ARG) is a rainfall observation instrument that can accurately measure rainfall at observation points. However, it has problems such as communication systems that cause delays in data transmission, low instrument density, and inability to cover a wide spatial area, which can affect the accuracy of rainfall information. Weather radar is a remote sensing instrument that can estimate rainfall spatially so that weather radar observations can reach areas of the region that do not have ARG. However, before being used as rainfall information, estimation rainfall needs to be evaluated or calibrated. Evaluation of rainfall estimation on weather radar to ARG in Banten at a 30– 120 km distance range, shows a coefficient of determination above 0.8. Based on the studies that have been conducted, increase of root mean square error (RMSE) is due to influence of radar observation range and observation distance on ARG. Adjustment of rainfall estimation improves the accuracy of rainfall estimation. Adjusting rainfall estimation can reduce RMSE by 50%.

Measuring Amazon Rainfall Intensity With Sound Recorders

AbstractGround weather observations are scarce in many parts of the globe, hampering effective climate monitoring and disaster management. In the Amazon basin, this occurs due to its remoteness and the challenging measurement of rainfall within the forest. Innovative rainfall estimation methods are thus requested to fill this gap. Here we present an approach to estimate rainfall based on sound measurements. We identified the best frequency range to estimate rainfall occurrence and intensity, trained classification and regression models with sound and rain gauge data collected in the Central Amazon during 9 months. By training a random forest classifier/regression model based on power spectrum values it was possible to identify and satisfactorily estimate hourly rainfall rates in two vegetation environments distinct from the training site, located 30 km from it. The proposed method is a promising approach for future weather monitoring in remote tropical areas.

Performance of the Climate Hazards Group InfraRed Precipitation with Station data (CHIRPS) Model in the Northeast Region of Brazil

Study is presented on the potential for using the tools contained in the Google Earth Engine (GEE) geospatial data processing and analysis platform in climatological studies in the Northeast Region of Brazil (NEB), in particular the validation of estimates of total annual rainfall obtained by the Climate Hazards Group InfraRed Precipitation with Station data (CHIRPS) model in seven homogeneous regions of the NEB. This data is made up of various sources of meteorological information, including information from satellites, numerical models and observations at weather stations in operational and research centers across the globe. The performance of the CHIRPS model was assessed using climatological data obtained from the National Meteorological Institute's (INMET) meteorological database, metrics and statistical tests. The results showed that the CHIRPS model has potential for use in six of the seven regions studied, with low errors and strong statistical correlation, and limitations in one of these regions. The conclusion is that this technological tool, which has a low operating cost and is free of charge, can be used to assist in climatological studies, planning and the development of strategies for mitigation, adaptation and coexistence with the dry periods observed in the sub-regions of the NEB, especially in regions that are in a situation of socio-economic vulnerability, as is the case in the semi-arid region of the NEB.

Spatiotemporal performance evaluation of high-resolution multiple satellite and reanalysis precipitation products over the semiarid region of India.

The present investigation evaluates three satellite precipitation products (SPPs), Multi-Source Weighted-Ensemble Precipitation (MSWEP), Global Precipitation Climatology Centre (GPCC), Climate Hazard Infrared Precipitation with Station Data (CHIRPS), and two reanalysis datasets, namely, the ERA5 atmosphere reanalysis dataset (ERA5) and Indian Monsoon Data Assimilation and Analysis (IMDAA), against the good quality gridded reference dataset (1991-2022) developed by the India Meteorological Department (IMD). The evaluation was carried out in terms of the rainfall detection ability and estimation accuracy of the products using metrics such as the false alarm ratio (FAR), probability of detection (POD), misses, root mean square error (RMSE), and percent bias (PBIAS). Among all the rainfall products, ERA5 had the best ability to capture rainfall events with a higher POD, followed by MSWEP. Both MSWEP and ERA5 had PODs of 70-100% in more than 90% of the grids and less than 35% of missing rainfall events in the entire Tamil Nadu. In the case of the rainfall estimation accuracy evaluation, the MSWEP exhibited superior performance, with lower RMSEs and biases ranging from - 25 to 25% at the annual and seasonal scales. In northeast monsoon (NEM), CHIRPS demonstrated a comparable performance to that of MSWEP in terms of the RMSE and PBIAS. These findings will help product users select the best reliable rainfall dataset for improved research, diversified applications in various sectors, and policy-making decisions.

Evaluation and inter-comparison of twenty-three gridded rainfall products representing a typical urban monsoon climate in India

Accurate and reliable estimation of rainfall is crucial for scientific research and various applications. However, the observed rainfall data is often limited. With the advancements in technology, many global gridded rainfall products are now available, but their accuracy levels vary across the world. In this study, we comprehensively analyzed the reliability and effectiveness of 23 publicly available global rainfall datasets against the observed rainfall for Patna, representing a typical urban monsoon climate in India. Thirteen continuous and ten categorical statistical metrics were applied at daily, weekly, monthly, and annual intervals over 16 years (2000–2015). The results indicate that the reliability of all derived rainfall datasets varied on different temporal scales and reference datasets used. Overall, in continuous metrics, MERRA2 and MSWEP consistently outperformed in all the temporal scales whereas in categorical metrics for analyzing the rainfall detection ability, AIMERG, followed by MERRA2 demonstrated superior performance among others. Furthermore, IMD GRID, GSMAP, PCCS, AIMERG, and IMERG performed well in estimating different rainfall intensities. MERRA2 and MSWEP, which have not been widely considered for evaluation in a monsoon climate were found to be outstanding performers consistently. Therefore, we suggest broadening the selection of global rainfall products in the evaluation to fully utilize the potentiality of all available options. Furthermore, our approach offers a reliable framework to comprehensively assess the performance of different gridded rainfall products and assist in the selection of the best rainfall product for a particular region and purpose.

Coupled Hydrogeophysical Modeling to Constrain Unsaturated Soil Parameters for a Slow‐Moving Landslide

AbstractGeophysical methods have proven to be useful for investigating unstable slopes as they are both non‐invasive and sensitive to the spatial distribution of physical properties in the subsurface. Of particular interest are the links between electrical resistivity and near‐surface moisture content; recent work has demonstrated that it is possible to calibrate hydrological models using geophysical measurements. In this study we explore the use of in‐field electrical resistivity data for calibrating unsaturated soil retention parameters and saturated hydraulic conductivity used for modeling unsaturated fluid flow. We study a synthetic case study, and a well‐characterized site in the northeast of England and develop an approach to calibrate retention parameters for a mudstone and a sandstone formation, the former being an actively failing unit. Petrophysical relationships between electrical resistivity and moisture content (or saturation) are established for both formations. 2D hydrological models are driven by effective rainfall estimations; subsequently these models are coupled with a geophysical forward model via a Markov chain Monte Carlo approach. For the synthetic case, we show that our modeling approach is sensitive to the moisture retention parameters, while less so to saturated hydraulic conductivity. We observe the same characteristics and sensitivities for the field case, albeit with a greater data misfit. Further hydrological simulations suggest that the slope retained high moisture contents in the months preceding a rotational failure. Therefore, we propose that coupled hydrological and geophysical modeling approaches could aid in enhancing landslide monitoring, modeling, and early warning efforts.

Quality Assessment of Satellite Data for Rainfall Estimation using Streamflow Simulation in Hydrological Modeling (Case Study: the Tighsiah Catchment)

Quality Assessment of Satellite Data for Rainfall Estimation using Streamflow Simulation in Hydrological Modeling (Case Study: the Tighsiah Catchment)