Search

Ground 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. Plain Language Summary: Understanding and predicting rainfall is a complex task, especially in areas where the availability of data from surface stations is limited, a common feature in many developing regions with insufficient rain gauge coverage. Recently, new opportunistic methods of rainfall measurement have emerged. Among them, is the use of the relationship between rainfall intensity and the sound produced by droplets hitting a surface. Sound recorders offer a low‐cost solution and could provide an interesting means to increase spatial coverage of rainfall measurements, but also to fill information gaps under dense forests where conventional devices do not work. Our study developed a new technique and applied it to the Central Amazon region, by training a supervised machine learning model applied to sound recordings obtained in a tropical rainforest. To our knowledge, for the first time, such techniques are validated in locations far from the calibration site. We showed that reasonable results can be obtained for sites with distinct vegetation types and up to 30 km of distance from where the training data was acquired. Our findings demonstrate a strong capability for estimating hourly rainfall rates. Key Points: Rainfall intensity estimated from sound measurements in the Amazon rainforest, tipping bucket rain gauge, and machine learning modelsThe best model successfully detects rainfall in 88% of the cases, with R2 > 0.87 for hourly rainfall rates on the training siteModel validated over two sites in the Amazon, 97% accuracy identifying rainfall events, R2 of 0.69 and 0.93 for hourly rainfall rate [ABSTRACT FROM AUTHOR]

This paper proposes a novel method to estimate rainfall intensity by analyzing the sound of raindrops. An innovative device for collecting acoustic data was designed, capable of blocking ambient noise in rainy environments. The device was deployed in real rainfall conditions during both the monsoon season and non-monsoon season to record raindrop sounds. The collected raindrop sounds were divided into 1 s, 10 s, and 1 min intervals, and the performance of rainfall intensity estimation for each segment length was compared. First, the rainfall occurrence was determined based on four extracted frequency domain features (average of dB, frequency-weighted average of dB, standard deviation of dB, and highest frequency), followed by a quantitative estimation of the rainfall intensity for the periods in which rainfall occurred. The results indicated that the best estimation performance was achieved when using 10 s segments, corresponding to the following metrics: accuracy: 0.909, false alarm ratio: 0.099, critical success index: 0.753, precision: 0.901, recall: 0.821, and F1 score: 0.859 for rainfall occurrence classification; and root mean square error: 1.675 mm/h, R 2 : 0.798, and mean absolute error: 0.493 mm/h for quantitative rainfall intensity estimation. The proposed small and lightweight device is convenient to install and manage and is remarkably cost-effective compared with traditional rainfall observation equipment. Additionally, this compact rainfall acoustic collection device can facilitate the collection of detailed rainfall information over vast areas. [ABSTRACT FROM AUTHOR]

Abstract Ground 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.

Accurate rainfall information is crucial for various applications, including river flow estimation, water resource management, and flood warning system development. Traditional rain gauge networks, however, suffer from limited spatial coverage, leading to incomplete and biased data for large areas. This study proposes a novel approach for surface rainfall estimation using weather radar data and a MultiLayer Perceptron (MLP) Regressor machine learning model. Grid search was employed to explore model performance across different windowing configurations: no windowing, n-1 windowing, and n-2 windowing. The results demonstrate that n-1 windowing outperforms other configurations, achieving an average RMSE of 0.987, MAE of 0.263, and R-squared of 0.242 across five locations. This suggests that n-1 windowing effectively captures the temporal dynamics of rainfall patterns while improving the model's sensitivity to regularization. However, a tendency for underestimating high-intensity rainfall events remains. This research highlights the effectiveness of n-1 windowing with MLP Regressors for enhanced surface rainfall estimation using weather radar data. Further investigation is needed to address the underestimation bias, particularly for high rainfall events.

Copyright of Journal of Education & Science is the property of Republic of Iraq Ministry of Higher Education & Scientific Research (MOHESR) and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Rainfall data is pivotal for hydrological studies, water resource management, and climate analysis. However, regions like Indonesia face challenges due to uneven rainfall station distribution. This study explores the potential of Tropical Rainfall Measuring Mission (TRMM) 3B42RT Daily V7 satellite data for rainfall estimation in such areas, focusing on the Brantas sub-watersheds: Kedak, Ngampel, and Kresek. Multiple TRMM scenarios are assessed for reliability and accuracy against ground-based rainfall station data. In monthly rainfall analysis, TRMM exhibits stronger correlations with average ground station data (coefficients of determination ranging from 0.62 to 0.74 for individual stations and 0.76 to 0.79 for averages) with an average PBIAS of 35.07%. The TRMM data exhibited weak performance on daily rainfall records, as evidenced by the low coefficient of determination (R²) values ranging from 0.04 to 0.21. TRMM proves sufficiently reliable for areas with limited rainfall records, particularly for cumulative monthly rainfall assessment. A comparative analysis of rainfall values for different return periods reveals that TRMM tends to provide lower values compared to single-station ground-based estimates. However, nearly similar values were obtained for area-averaged ground-based stations. Combining area-averaged TRMM data with cumulative ground-based rainfall data reduces variability in values for diverse return periods (3-8%), enhancing data accuracy. In conclusion, TRMM 3B42RT Daily V7 data integration enhances rainfall estimation accuracy in regions with limited data availability. [ABSTRACT FROM AUTHOR]

Rainfall is a crucial component of the hydrological cycle and plays a key role in water resource planning. Recent research has investigated the use of gridded data as a supplement to and replacement for traditional rain gauge measurements, particularly in areas with limited gauge coverage. Gridded precipitation data offering a structured method to represent precipitation patterns across large regions by dividing the data into grids. This enables more precise spatial analysis of precipitation distribution and variability. The study assessed the accuracy of six high-resolution gridded rainfall product estimates (ERA5, ERA-Interim, CMORPH, PERSIANN, PERSIANN-CDR, and PERSIANN-CCS) at 12 rain gauge stations in Iran at various time scales. Comparisons with rain gauge network data using statistical and graphical methods revealed that ERA5, ERA-Interim, and PERSIANN-CDR data outperformed other models on annual and monthly scales, so that the highest correlation coefficient in monthly scale was obtained by ERA5 model at Doroodzan station with correlation coefficient of 0.93. Also, the results on a daily scale indicate the appropriateness of the output data of the reanalysis models (ERA5, ERA-Interim) compared to other models in such a way that the lowest RMSE value in all stations except Sefidroud Dam is related to the reanalysis data and the lowest RMSE value is equal to with 0.78 mm at the Chahnimeh station and the highest value of the correlation coefficient equal to 0.63 corresponds to the Karaj dam rain gauge station; Also, in correctly detecting rainy and non-rainy days, ERA5 model has the most accuracy in all stations.

Introduction and GoalUsers prepare accurate data of the amount of rainfall by using rain gauge stations. However, an interpolation of rainfall data is difficult due to temporal and spatial variability. Therefore, rain gauge stations are not well distributed in many areas, especially in mountainous areas. In a mountainous area, understanding the interaction between the resolution of the Digital Elevation Model (DEM) and climate variables is necessary for accurate spatial interpolation of average rainfall in many areas, and on the other hand, the need for accurate information in hydrological modeling and many environmental studies and it is climatic. One of the problems that exists in many hydrological studies is that rainfall maps are always prepared using interpolation or available DEM, regardless of rainfall, which have an estimated rainfall error.Materials and MethodsIn this study, four DEMs with spatial resolutions of 30, 90, 1000, and 10000 m, which are the most common DEMs in studies, were used to introduce the best elevation digital model for extracting the rainfall gradient map from the data of 11 meteorological stations in Kermanshah province. A rainfall map for Kermanshah province was prepared using a linear regression model fitted between the height of each station and the 20-year average rainfall. The best DEM for rainfall estimation was then determined on the basis of error evaluation criteria.ResultsThe results of this research showed that in estimating rainfall, DEMs with cell sizes of 1000 and 10000 m (R2 = 0.76, 0.81) were more accurate than DEMs with spatial accuracy of 30 and 90 m (R2 = 0.75). In the examination of the Nash–Sutcliffe coefficient (NS), compared to other digital height models of accuracy, DEM with a spatial resolution of 1000 m (one km) with a Nash–Sutcliffe coefficient of 0.76, a significance level of 0.01, and a correlation coefficient of 0.81 was found to have greater accuracy.Conclusion and SuggestionsThe results of the present study can be used to estimate and generalize rainfall in areas that do not have stations and to prepare rainfall maps in areas where the number of stations is limited. In addition, it should be used in univariate interpolation methods that do not have proper accuracy because spatial distances are not considered. In addition, due to the complex topography of the earth and the non-uniformity of meteorological stations on the earth’s surface, high-resolution models with higher spatial resolution are required for the estimation of rainfall, which increases the accuracy of digital models in the evaluation of rainfall studies by removing topographical levels that cause errors.

The analysis of the spatial and temporal distribution of storm events contributes to a better use of water resources, for example, the supply of drinking water, irrigation practices, electricity generation and management of extreme events to control floods and mitigate droughts, among others. The traditional observation of rainfall fields in Mexico has been carried out using rain gauge network data, but their spatial representativeness is unsatisfactory. Therefore, this study reviewed the possibility of obtaining better estimates of the spatial distribution of daily rainfall considering information from three different databases, which include rain gauge measurements and remotely sensed precipitation products of satellite systems and weather radars. In order to determine a two-dimensional rainfall distribution, the information has been merged with a sequential data assimilation scheme up to the diagnostic stage, paying attention to the benefit that the rain gauge network density has on the estimation. With the application of the Barnes method, historical events in the Mexican territory were analyzed using statistical parameters for the validation of the estimates, with satisfactory results because the assimilated rainfalls turned out to be better approximations than the values calculated with the individual databases, even for a not very low density of surface observations. HIGHLIGHTS The merging of the three databases considered allows for determining satisfactory approximations of the rainfall fields analyzed, regardless the rain gauge network configuration.; The increment in the error rates of the assimilated rainfall estimates is mainly due to the lack of accuracy of the remote sensing products.; The recovery of the spatial behavior of historical storms is possible.;

This paper proposes a novel method to estimate rainfall intensity by analyzing the sound of raindrops. An innovative device for collecting acoustic data was designed, capable of blocking ambient noise in rainy environments. The device was deployed in real rainfall conditions during both the monsoon season and non-monsoon season to record raindrop sounds. The collected raindrop sounds were divided into 1 s, 10 s, and 1 min intervals, and the performance of rainfall intensity estimation for each segment length was compared. First, the rainfall occurrence was determined based on four extracted frequency domain features (average of dB, frequency-weighted average of dB, standard deviation of dB, and highest frequency), followed by a quantitative estimation of the rainfall intensity for the periods in which rainfall occurred. The results indicated that the best estimation performance was achieved when using 10 s segments, corresponding to the following metrics: accuracy: 0.909, false alarm ratio: 0.099, critical success index: 0.753, precision: 0.901, recall: 0.821, and F1 score: 0.859 for rainfall occurrence classification; and root mean square error: 1.675 mm/h, R2: 0.798, and mean absolute error: 0.493 mm/h for quantitative rainfall intensity estimation. The proposed small and lightweight device is convenient to install and manage and is remarkably cost-effective compared with traditional rainfall observation equipment. Additionally, this compact rainfall acoustic collection device can facilitate the collection of detailed rainfall information over vast areas.

As rainfall exhibits high spatiotemporal variability, accurate and real-time rainfall monitoring is vitally important in fields such as hydrometeorological research, agriculture and disaster prevention and control. Nevertheless, the current dedicated rain sensors cannot fulfill the requirement for comprehensive precipitation observation, owing to their respective limitations. Within the last two decades, the utilization of commercial microwave links (CMLs) for rainfall estimation, as an opportunistic sensing method, has generated considerable attention. Relying on CML networks deployed and maintained by mobile network operators can provide near-surface precipitation information over large areas at a low cost. Although scholars have developed several algorithms for obtaining rainfall estimates from CML data, the rainfall estimation technique based on CMLs remains challenging due to the complex effect in the microwave radiation transmission process. In this paper, we provide a comprehensive review of the technical principles, developments and workflows for this technology, alongside its application in environmental monitoring and hydrological modeling. Furthermore, this paper outlines the current challenges and future research directions, which will hopefully draw the attention of researchers and provide valuable guidance. [ABSTRACT FROM AUTHOR]

Devastating floods have plagued many West African cities in the past decades. In an attempt to reduce flood damage in Bamako (Mali), an early warning system (EWS) demonstrator (Raincell App) was developed for flash floods. On 16 May 2019, while the demonstrator was partially operational, an intense rainfall event led to devastating floods. We carried out an experience feedback on this flood event by comparing EWS simulations to the results of a field survey. Given the synoptic situation and the rapid development pattern of the storm, none of the global forecasting systems were able to foresee its occurrence and magnitude. The hydrological model developed as part of the demonstrator correctly identified most of the locations where overbank flow occurred. In the absence of data, the predicted discharge and volume values could not be validated. However, they are realistic based on the water levels reported in the Post‐Disaster Needs Assessment report. It would be advisable to couple it to a two‐dimensional hydraulic model and add discharge and water level monitoring to the already existing rainfall surveillance scheme to further improve the system's performance. Increasing the local population's awareness of the dangers of clogged waterways is also mandatory. [ABSTRACT FROM AUTHOR]

The SM2RAIN (soil moisture to rain) model has been widely used for rainfall estimation worldwide. However, due to the lack of sufficient ground observation, the SM2RAIN model driven by different passive microwave soil moisture products over the Tibetan Plateau has not been fully validated. In this article, four widely used satellite microwave soil moisture products (including SMAP, ASCAT, SMOS, and AMSR2) were used as input data for rainfall estimation. Rainfall data from eight ground observation stations during 2016–2018 were used to evaluate the overall performance of the SM2RAIN algorithm under various soil moisture products at different time aggregation scales. In addition, different satellite soil moisture products were merged to evaluate whether the combined soil moisture products could improve the performance of the SM2RAIN model. Finally, the rainfall estimates with different soil moisture data were further evaluated and compared with two benchmark rainfall products (IMERG and ERA5). Results indicate that: 1) Overall, SM2RAIN-SMAP has the highest rainfall estimation accuracy, but with the time aggregation scale up to 30 days, the mean R of the four rainfall estimates could reach above 0.8 and the mean value of Kling–Gupta efficiency could reach above 0.8. 2) Combined satellite soil moisture products can significantly improve the rainfall estimates. The SM2RAIN model performed the best when SMAP and ASCAT soil moisture products were combined. 3) Using the SMAP product or combined soil moisture products yielded more accurate rainfall estimates than the two benchmark rainfall products (IMERG and ERA5).

Accurate rainfall estimation is crucial to adequately assess the risk associated with extreme events capable of triggering floods and landslides. Data gathered from Rain Gauges (RGs), sensors devoted to measuring the intensity of the rain at individual points, are commonly used to feed interpolation methods (e.g., the Kriging geostatistical approach) and estimate the precipitation field over an area of interest. However, the information provided by RGs could be insufficient to model complex phenomena, and computationally expensive interpolation methods could not be used in real-time environments. Integrating additional data sources (e.g., radar and geostationary satellites) is an effective solution for improving the quality of the estimate, but it needs to cope with Big Data issues. To overcome all these issues, we propose a Rainfall Estimation Model (REM) based on an Ensemble of Deep Neural Networks (DeepEns-REM) that can automatically fuse heterogeneous data sources. The usage of Residual Blocks in the base models and the adoption of a Snapshot procedure to build the ensemble guarantees a fast convergence and scalability. Experimental results, conducted on a real dataset concerning a southern region in Italy, demonstrate the quality of the proposal in comparison with the Kriging interpolation technique and other machine learning techniques, especially in the case of exceptional rainfall events. [ABSTRACT FROM AUTHOR]

The evaluation of the effects of artificial precipitation enhancement remains one of the most important and challenging issues in the fields of meteorology. Rainfall is the most important evaluation metric for artificial precipitation enhancement, which is mainly achieved through physics-based models that simulate physical phenomena and data-driven statistical models. The series of effect evaluation methods requires the selection of a comparison area for effect comparison, and idealized assumptions and simplifications have been made for the actual cloud precipitation process, leading to unreliable quantitative evaluation results of artificial precipitation effects. This paper proposes a deep learning-based method (UNET-GRU) to quantitatively evaluate the effect of artificial rainfall. By comparing the residual values obtained from inverting the natural evolution grid rainfall of the same area under the same artificial rainfall conditions with the actual rainfall amount after artificial rainfall operations, the effect of artificial rainfall can be quantitatively evaluated, effectively solving the problem of quantitative evaluation of artificial precipitation effects. Wuhan and Shiyan in China are selected to represent typical plains and mountainous areas, respectively, and the method is evaluated using 6-min resolution radar weather data from 2017 to 2020. During the experiment, we utilized the UNET-GRU algorithm and developed separate algorithms for comparison against common persistent baselines (i.e., the next-time data of the training data). The prediction of mean squared error (MSE) for these three algorithms was significantly lower than that of the baseline data. Moreover, the indicators for these algorithms were excellent, further demonstrating their efficacy. In addition, the residual results of the estimated 7-h grid rainfall were compared with the actual recorded rainfall to evaluate the effectiveness of artificial precipitation. The results showed that the estimated rainfall was consistent with the recorded precipitation for that year, indicating that deep learning methods can be successfully used to evaluate the impact of artificial precipitation. The results demonstrate that this method improves the accuracy of effect evaluation and enhances the generalization ability of the evaluation scheme. [ABSTRACT FROM AUTHOR]

As rainfall exhibits high spatiotemporal variability, accurate and real-time rainfall monitoring is vitally important in fields such as hydrometeorological research, agriculture and disaster prevention and control. Nevertheless, the current dedicated rain sensors cannot fulfill the requirement for comprehensive precipitation observation, owing to their respective limitations. Within the last two decades, the utilization of commercial microwave links (CMLs) for rainfall estimation, as an opportunistic sensing method, has generated considerable attention. Relying on CML networks deployed and maintained by mobile network operators can provide near-surface precipitation information over large areas at a low cost. Although scholars have developed several algorithms for obtaining rainfall estimates from CML data, the rainfall estimation technique based on CMLs remains challenging due to the complex effect in the microwave radiation transmission process. In this paper, we provide a comprehensive review of the technical principles, developments and workflows for this technology, alongside its application in environmental monitoring and hydrological modeling. Furthermore, this paper outlines the current challenges and future research directions, which will hopefully draw the attention of researchers and provide valuable guidance.

Assessing the uncertainty of precipitation measurements is a challenging problem because precipitation estimates are inevitably influenced by various errors and environmental conditions. A way to characterize the error structure of coincident measurements is to use the triple colocation (TC) statistical method. Unlike more typical approaches, where measures are compared in pairs and one of the two is assumed error‐free, TC has the enviable advantage to succeed in characterizing the uncertainties of co‐located measurements being compared to each other, without requiring the knowledge of the true value which is often unknown. However, TC requires to have at least three co‐located measuring systems and the compliance with several initial assumptions. In this work, for the first time, TC is applied to in‐situ measurements of rain precipitation acquired by three co‐located devices: a weighing rain gauge, a laser disdrometer and a bidimensional video disdrometer. Both parametric and nonparametric formulations of TC are implemented to derive the rainfall product precision associated with the three devices. While the parametric TC technique requires tighter constraints and explicit assumptions which may be violated causing some artifacts, the nonparametric formulation is more flexible and requires less strict constrains. For this reason, a comparison between the two TC formulations is also presented to investigate the impact of TC constrains and their possible violations. The results are obtained using a statistically robust dataset spanning a 1.5 year period collected in Switzerland and presented in terms of traditional metrics. According to triple colocation analysis, the two disdrometers outperform the classical weighing rain gauge and they have similar measurement error structure regardless of the integration time intervals. [ABSTRACT FROM AUTHOR]

Recently, there have been many abnormal natural phenomena caused by climate change. Anthropogenic factors associated with insufficient water resource management can be another cause. Among natural causes, rainfall intensity and volume often induce flooding. Therefore, accurate rainfall estimation and prediction can prevent and mitigate damage caused by these hazards. Sadly, uncertainties often hinder accurate rainfall forecasting. This study investigates the uncertainty of the Korean rainfall ensemble prediction data and runoff analysis model in order to enhance reliability and improve prediction. The objectives of this study include: (i) evaluating the spatial characteristics and applicability of limited area ensemble prediction system (LENS) data; (ii) understanding uncertainty using parameter correction and generalized likelihood uncertainty estimation (GLUE) and grid-based rainfall-runoff model (GRM); (iii) evaluating models before and after LENS-GRM correction. In this study, data from the Wicheon Basin was used. The informal likelihood (R2, NSE, PBIAS) and formal likelihood (log-normal) were used to evaluate model applicability. The results confirmed that uncertainty of the behavioral model exists using the likelihood threshold when applying the runoff model to rainfall forecasting data. Accordingly, this method is expected to enable more reliable flood prediction by reducing the uncertainties of the rainfall ensemble data and the runoff model when selecting the behavioral model for the user's uncertainty analysis. It also provides a basis for flood prediction studies that apply rainfall and geographical characteristics for rainfall-runoff uncertainty analysis. [ABSTRACT FROM AUTHOR]

Rainfall data have a profound significance for meteorology, climatology, hydrology, and environmental sciences. However, existing rainfall observation methods (including ground-based rain gauges and radar-/satellite-based remote sensing) are not efficient in terms of spatiotemporal resolution and cannot meet the needs of high-resolution application scenarios (urban waterlogging, emergency rescue, etc.). Widespread surveillance cameras have been regarded as alternative rain gauges in existing studies. Surveillance audio, through exploiting their nonstop use to record rainfall acoustic signals, should be considered a type of data source to obtain high-resolution and all-weather data. In this study, a method named parallel neural network based on attention mechanisms and multi-scale fusion (PNNAMMS) is proposed for automatically classifying rainfall levels by surveillance audio. The proposed model employs a parallel dual-channel network with spatial channel extracting the frequency domain correlation, and temporal channel capturing the time-domain continuity of the rainfall sound. Additionally, attention mechanisms are used on the two channels to obtain significant spatiotemporal elements. A multi-scale fusion method was adopted to fuse different scale features in the spatial channel for more robust performance in complex surveillance scenarios. In experiments showed that our method achieved an estimation accuracy of 84.64% for rainfall levels and outperformed previously proposed methods. [ABSTRACT FROM AUTHOR]

Weather radar research has produced numerous radar-based rainfall estimators based on climate, rainfall intensity, a variety of ground-truthing instruments and sensors (e.g., rain gauges, disdrometers), and techniques. Although each research direction gives improvement, their collective application in an operational sense still yields uncertainty in rainfall estimation at times. This study aims to explore the concept of implementing Machine Learning (ML) models in optimizing the radar-based rainfall estimations at the bin level from a group of estimator. The Canadian King City C-Band radar was used with a GEONOR T-200B rain gauge (a total of 263 sample points) to establish a group of polarimetric-based rainfall estimators (R(Z), R(Z, ZDR), R(KDP)). The estimators were used to train three ML models, namely Decision Tree, Random Forest, and Gradient Boost, to choose the optimal rainfall estimators based on radar variables (Z, ZDR, KDP). Data from the Canadian Exeter C-Band radar and a Texas Electronics TE525 tipping bucket gauge at a different location were used to verify the ML models and compare their results to the most commonly used Z-R relations. The verification process shows promising results for the ML models, specifically the Gradient Boost model. These encouraging results need to be further explored with more sample points to further refine the ML models. [ABSTRACT FROM AUTHOR]

This study of the Quantitative Estimation Precipitation (QEP) of rainfall, detected by two Meteorology Radars over Chi Basin, North-east Thailand, used data from the Thai Meteorological Department (TMD). The rainfall data from 129 rain gauge stations in the Chi Basin area, covering a period of two years, was also used. The study methodology consists of: firstly, deriving the QPE between radar and rainfall based on meteorological observations using the Marshall Palmer Stratiform, the Summer Deep Convection, and Regression Model and calibrating with rain gauge station data; secondly, Bias Correction using statistical method; thirdly, determining spatial variation using three methods, namely Kriging, Inverse Distance Weight (IDW), and the Minimum Curvature Method. The results of the study demonstrated the accuracy of estimating precipitation using meteorological radar. Estimated precipitation compared against an equivalent of 2 years of rain station measurement had a probability of detection (POD) of 0.927, where a value of 1 indicated perfect agreement, demonstrating the effectiveness of the method used to calibrate the radar data. The bias correction method gave high accuracy compared with measured rainfall. Furthermore, of the spatial estimation of rainfall methods, the Kriging methodology showed the best fit between estimation of rainfall distribution and measured rainfall distribution. Therefore, the results of this study showed that the rainfall estimation, using data from a meteorology radar, has good accuracy and can be useful, especially in areas where it is not possible to install and operate rainfall measurement stations, such as in heavily forested areas and/or in steep terrain. Additionally, good accuracy rainfall data derived from radar data can be integrated with other data used for water management and natural disasters for applications to reduce economic losses, as well as losses of life and property. [ABSTRACT FROM AUTHOR]

Abstract Assessing the uncertainty of precipitation measurements is a challenging problem because precipitation estimates are inevitably influenced by various errors and environmental conditions. A way to characterize the error structure of coincident measurements is to use the triple colocation (TC) statistical method. Unlike more typical approaches, where measures are compared in pairs and one of the two is assumed error‐free, TC has the enviable advantage to succeed in characterizing the uncertainties of co‐located measurements being compared to each other, without requiring the knowledge of the true value which is often unknown. However, TC requires to have at least three co‐located measuring systems and the compliance with several initial assumptions. In this work, for the first time, TC is applied to in‐situ measurements of rain precipitation acquired by three co‐located devices: a weighing rain gauge, a laser disdrometer and a bidimensional video disdrometer. Both parametric and nonparametric formulations of TC are implemented to derive the rainfall product precision associated with the three devices. While the parametric TC technique requires tighter constraints and explicit assumptions which may be violated causing some artifacts, the nonparametric formulation is more flexible and requires less strict constrains. For this reason, a comparison between the two TC formulations is also presented to investigate the impact of TC constrains and their possible violations. The results are obtained using a statistically robust dataset spanning a 1.5 year period collected in Switzerland and presented in terms of traditional metrics. According to triple colocation analysis, the two disdrometers outperform the classical weighing rain gauge and they have similar measurement error structure regardless of the integration time intervals.

High spatio-temporal resolution and accurate long-term rainfall estimates are critical in sustainable water resource planning and management, assessment of climate variability and extremes, and hydro-meteorology-related water system decisions. The recent advent of improved higher-resolution open-access satellite-based rainfall products has emerged as a viable complementary to ground-based observations that can often not capture the rainfall variability on a spatial scale. In a developing country such as Nepal, where the rain-gauge monitoring network is sparse and unevenly distributed, satellite rainfall estimates are crucial. However, substantial errors associated with such satellite rainfall estimates pose a challenge to their application, particularly in complex orographic regions such as Nepal. Therefore, these precipitation products must be validated before practical usage to check their accuracy and occurrence consistency. This study aims to assess the reliability of the Climate Hazards Group Infrared Precipitation with Stations (CHIRPS) product against ground-based observations from 1986 to 2015 in five medium-sized river basins in Nepal, namely, Babai, Bagmati, Kamala, Kankai, and the West Rapti river basin. A set of continuous evaluation metrics (correlation coefficient, root mean square error, relative bias, and Kling-Gupta efficiency) were used in analyzing the accuracy of CHIRPS and categorical metrics (probability of detection, critical success index, false alarm ratio, and frequency bias index). The Probability of Detection and Critical Success Index values were found to be considerably low (<0.4 on average), while the false alarm ratio was significant (>0.4 on average). It was found that CHIRPS showed better performance in seasonal and monthly time scales with high correlation and indicated greater consistency in non-monsoon seasons. Rainfall amount (less than 10 mm and greater than 150 mm) and rainfall frequency was underestimated by CHIRPS in all basins, while the overestimated rainfall was between 10 and 100 mm in all basins except Kamala. Additionally, CHIRPS overestimated dry days and maximum consecutive dry days in the study area. Our study suggests that CHIRPS rainfall products cannot supplant the ground-based observations but complement rain-gauge networks. However, the reliability of this product in capturing local extreme events (such as floods and droughts) seems less prominent. A high-quality rain gauge network is essential to enhance the accuracy of satellite estimations. [ABSTRACT FROM AUTHOR]

The increased frequency of flooding in the metropolitan area of São Paulo (MASP), São Paulo, Brazil, makes it necessary to have an accurate precipitation monitoring and nowcasting system to mitigate socioeconomic impacts. This research is on the use of remote sensing with dual Doppler X-band weather radar (MXPOL) to improve the high spatiotemporal resolution rainfall rates estimation in MASP. The methodology includes an initial correction of the polarimetric data, namely, the specific differential phase (KDP) to estimate the attenuation correction, the adjustment of rainfall rates by polarimetric variables relationships, and the error estimation by cross-validation with rain gauges. Results show that the five relationships have similar performance with just a slight difference between the use of power relationships (e.g., reflectivity (Z) and differential reflectivity (ZDR)) or frequency (e.g., specific differential phase) variables. The average error was 1.1 mm for the mean square error, 29% for the relative error, and 0.8 for the correlation coefficient. An adjustment of coefficient specifically for MXPOL was significantly advantageous for the rainfall rate estimation with R (Z, ZDR) and R (ZDR, KDP) relationships. [ABSTRACT FROM AUTHOR]

Nowadays, due to the high uncertainty in estimating precipitation in different geographical areas, the use of computational intelligence methods based on optimization algorithms to accurately estimate daily precipitation has been considered by water engineers. In the present study, the combined Adaptive Neuro Fuzzy Inference System and Wavelet transform (W-ANFIS) method was used as a pre-processor for daily rainfall data to estimate precipitation values. The structure of the W-ANFIS hybrid model was developed using the Fuzzy Clustering Means (FCM) method in the training phase. Moreover, constant coefficients of membership functions applied in the ANFIS model were optimized using four optimization algorithms including Genetic Algorithm (GA), Particle Swarm Optimization (PSO), Differential Evolution (DE), and Ant Colony community (ACO). In the present study, rainfall statistics of Izmir basin in the western part of Turkey were used. Through applying five-time delays in daily rainfall statistics as well as decomposing each time delay in the three levels of wavelet transform, each of the W-ANFIS optimal models had twenty input variables. The results of the statistical analysis for both training and testing stages by the Root Mean Square Error (RMSE) and Mean Absolute Error (MAE) showed that the application of the DE algorithm into W-ANFIS structure had the best performance (RMSE = 22.22 and MAE = 17.11mm) than other combined models with PSO (RMSE = 28.11 and MAE = 24.11 mm), ACO (RMSE = 30.41 and MAE = 26.50 mm), and GA (RMSE = 25.70 and MAE = 18.11 mm).

An approach to developing a blended satellite-rainfall dataset over Australia that could be suitable for operational use is presented. In this study, Global Satellite Mapping of Precipitation (GSMaP) satellite precipitation estimates were blended with station-based rain gauge data over Australia, using operational station data that has not been harnessed by other blended products. A two-step method was utilized. First, GSMaP satellite precipitation estimates were adjusted using rain gauge data through multiplicative ratios that were gridded using ordinary kriging. This step resulted in reducing dry biases, especially over topography. The adjusted GSMaP data was then blended with the Australian Gridded Climate Dataset (AGCD) rainfall analysis, an operational station-based gridded rain gauge dataset, using an inverse error variance weighting method to further remove biases. A validation that was performed using a 20-year range (2001 to 2020) showed the proposed approach was successful; the resulting blended dataset displayed superior performance compared to other non-gauge-based datasets with respect to stations as well as displaying more realistic patterns of rainfall than the AGCD in areas with no rain gauges. The average mean absolute error (MAE) against station data was reduced from 0.89 to 0.31. The greatest bias reductions were obtained for extreme precipitation totals and over mountainous regions, provided sufficient rain gauge availability. The newly produced dataset supported the identification of a general positive bias in the AGCD over the north-west interior of Australia. [ABSTRACT FROM AUTHOR]

The estimation of maximum rainfall at different return periods (T) receives a crucial role in the precise planning of irrigation systems, hydraulic structures, and drainage systems. The design of such structures should be done in such a way that these structures should not be damaged due to extreme rainfall events in their entire life. The probability/frequency analysis of maximum rainfall is necessary for the selection of an appropriate model that could anticipate extreme natural processes like rainfall and flood. This study aims to select the best probability distribution function (PDF) and extrapolate maximum rainfall magnitudes at different time scales for higher durations of T in the Parbhani district of Maharashtra state of India. The maximum rainfall analysis for the study area was accomplished using daily rainfall data of 49 years (1971 to 2019). Weibull's plotting position (WPP) formula was used to estimate observed rainfall magnitudes at different durations of T. Different PDFs like Gumbel extreme value (GEV), log-Pearson type III (LP-III), Pearson type III (P-III), log normal (L-NM), and normal (NM) were used at different time scales for estimation of annual, maximum monthly, and consecutive 1, 2, 3, 4, and 5 days of maximum rainfall. The estimated magnitudes of maximum rainfall were compared with the results of the WPP. The chi-square test was used for the selection of the best-fitting PDF for the study area. In this study, GEV PDF performed the best for annual, maximum monthly, and 1 day and consecutive 5 days of maximum rainfall estimation while LP-III PDF performed the best for consecutive 2, 3, and 4 days of maximum rainfall estimation of the study area. The values of maximum rainfall at different time scales were extrapolated for the 100- and 200-year durations of T using the selected PDF for the study area. These findings will help design engineers and planners in the construction of adequate drainage systems and hydraulic structures. The result of this study will be useful to formulate the appropriate strategy against flood hazards and damages. [ABSTRACT FROM AUTHOR]

Since the 1990s, mobile telecommunication networks have gradually become denser around the world. Nowadays, large parts of their backhaul network consist of commercial microwave links (CMLs). Since CML signals are attenuated by rainfall, the exploitation of records of this attenuation is an innovative and an inexpensive solution for precipitation monitoring purposes. Performance data from mobile operators’ networks are crucial for the implementation of this technology. Therefore, a real-time system for collecting and storing CML power levels from the mobile phone operator “Telecel Faso” in Burkina Faso has been implemented. This new acquisition system, which uses the Simple Network Management Protocol (SNMP), can simultaneously record the transmitted and received power levels from all the CMLs to which it has access, with a time resolution of one minute. Installed at “Laboratoire des Matériaux et Environnement de l’Université Joseph KI-ZERBO (Burkina Faso)”, this acquisition system is dynamic and has gradually grown from eight, in 2019, to more than 1000 radio links of Telecel Faso’s network in 2021. The system covers the capital Ouagadougou and the main cities of Burkina Faso (Bobo Dioulasso, Ouahigouya, Koudougou, and Kaya) as well as the axes connecting Ouagadougou to these cities.

Rainfall data have a profound significance for meteorology, climatology, hydrology, and environmental sciences. However, existing rainfall observation methods (including ground-based rain gauges and radar-/satellite-based remote sensing) are not efficient in terms of spatiotemporal resolution and cannot meet the needs of high-resolution application scenarios (urban waterlogging, emergency rescue, etc.). Widespread surveillance cameras have been regarded as alternative rain gauges in existing studies. Surveillance audio, through exploiting their nonstop use to record rainfall acoustic signals, should be considered a type of data source to obtain high-resolution and all-weather data. In this study, a method named parallel neural network based on attention mechanisms and multi-scale fusion (PNNAMMS) is proposed for automatically classifying rainfall levels by surveillance audio. The proposed model employs a parallel dual-channel network with spatial channel extracting the frequency domain correlation, and temporal channel capturing the time-domain continuity of the rainfall sound. Additionally, attention mechanisms are used on the two channels to obtain significant spatiotemporal elements. A multi-scale fusion method was adopted to fuse different scale features in the spatial channel for more robust performance in complex surveillance scenarios. In experiments showed that our method achieved an estimation accuracy of 84.64% for rainfall levels and outperformed previously proposed methods.

This study of the Quantitative Estimation Precipitation (QEP) of rainfall, detected by two Meteorology Radars over Chi Basin, North-east Thailand, used data from the Thai Meteorological Department (TMD). The rainfall data from 129 rain gauge stations in the Chi Basin area, covering a period of two years, was also used. The study methodology consists of: firstly, deriving the QPE between radar and rainfall based on meteorological observations using the Marshall Palmer Stratiform, the Summer Deep Convection, and Regression Model and calibrating with rain gauge station data; secondly, Bias Correction using statistical method; thirdly, determining spatial variation using three methods, namely Kriging, Inverse Distance Weight (IDW), and the Minimum Curvature Method. The results of the study demonstrated the accuracy of estimating precipitation using meteorological radar. Estimated precipitation compared against an equivalent of 2 years of rain station measurement had a probability of detection (POD) of 0.927, where a value of 1 indicated perfect agreement, demonstrating the effectiveness of the method used to calibrate the radar data. The bias correction method gave high accuracy compared with measured rainfall. Furthermore, of the spatial estimation of rainfall methods, the Kriging methodology showed the best fit between estimation of rainfall distribution and measured rainfall distribution. Therefore, the results of this study showed that the rainfall estimation, using data from a meteorology radar, has good accuracy and can be useful, especially in areas where it is not possible to install and operate rainfall measurement stations, such as in heavily forested areas and/or in steep terrain. Additionally, good accuracy rainfall data derived from radar data can be integrated with other data used for water management and natural disasters for applications to reduce economic losses, as well as losses of life and property.

Weather radar research has produced numerous radar-based rainfall estimators based on climate, rainfall intensity, a variety of ground-truthing instruments and sensors (e.g., rain gauges, disdrometers), and techniques. Although each research direction gives improvement, their collective application in an operational sense still yields uncertainty in rainfall estimation at times. This study aims to explore the concept of implementing Machine Learning (ML) models in optimizing the radar-based rainfall estimations at the bin level from a group of estimator. The Canadian King City C-Band radar was used with a GEONOR T-200B rain gauge (a total of 263 sample points) to establish a group of polarimetric-based rainfall estimators (R(Z), R(Z, ZDR), R(KDP)). The estimators were used to train three ML models, namely Decision Tree, Random Forest, and Gradient Boost, to choose the optimal rainfall estimators based on radar variables (Z, ZDR, KDP). Data from the Canadian Exeter C-Band radar and a Texas Electronics TE525 tipping bucket gauge at a different location were used to verify the ML models and compare their results to the most commonly used Z-R relations. The verification process shows promising results for the ML models, specifically the Gradient Boost model. These encouraging results need to be further explored with more sample points to further refine the ML models.

The amount of precipitation measured by the radar is different from the amount of precipitation received on the ground. This difference has many causes, some of which are related to the nature of the radar and others to the climate of each region. As a result, radar data needs to be corrected for radar data based on terrestrial data to determine the amount of ground-level rainfall received from the radar data. Weather radar used for estimation of rain in the large areas. The relationship between rain and reflectivity radar is exponential Z = aRb. Radar estimated rainfall amount is incorrect if the coefficients of this model are wrong. Drop size and distribution of rainfall is Effective on the coefficient of this model. The change in the coefficients of this model is very high. In this study, to calibrate radar data, rain from 2 to 3 December 2016 and 11 to 13 February 2018 at the stations, Kermanshah, Eslamabad, Sarpol, Ghasre Shirin, Harsin, Javanroud, Tazabad, Songhor, Ravansar, Ghilan Gharb and Soumar at distance of 30 to 100 kilometers from Kermanshah’s radar are investigated. In the first rain, using soft Rainbow for each of the stations and radar beam elevation angle optimization and correction factors relating to the extraction station, respectively. With this relationship, the radar rainfall estimates from 31 percent to 96 percent Increased and the average total rainfall from 8.9 to 32.4 millimeter increased an average radar rainfall estimated only 1 millimeter less than actual rain by gauge. In the second rain, using data from all stations, only one equation and correction factors were obtained. The results rainfall radar will be accepted at this stage, good approximation, and the average estimate rainfall radar from 9.6 to 23.5 millimeter increased those 4 millimeters less than the actual amount by gauge. If radar coefficients are corrected correctly for different areas, precipitation can be predicted and prevented from occurring unexpected events.

High spatio-temporal resolution and accurate long-term rainfall estimates are critical in sustainable water resource planning and management, assessment of climate variability and extremes, and hydro-meteorology-related water system decisions. The recent advent of improved higher-resolution open-access satellite-based rainfall products has emerged as a viable complementary to ground-based observations that can often not capture the rainfall variability on a spatial scale. In a developing country such as Nepal, where the rain-gauge monitoring network is sparse and unevenly distributed, satellite rainfall estimates are crucial. However, substantial errors associated with such satellite rainfall estimates pose a challenge to their application, particularly in complex orographic regions such as Nepal. Therefore, these precipitation products must be validated before practical usage to check their accuracy and occurrence consistency. This study aims to assess the reliability of the Climate Hazards Group Infrared Precipitation with Stations (CHIRPS) product against ground-based observations from 1986 to 2015 in five medium-sized river basins in Nepal, namely, Babai, Bagmati, Kamala, Kankai, and the West Rapti river basin. A set of continuous evaluation metrics (correlation coefficient, root mean square error, relative bias, and Kling-Gupta efficiency) were used in analyzing the accuracy of CHIRPS and categorical metrics (probability of detection, critical success index, false alarm ratio, and frequency bias index). The Probability of Detection and Critical Success Index values were found to be considerably low (0.4 on average). It was found that CHIRPS showed better performance in seasonal and monthly time scales with high correlation and indicated greater consistency in non-monsoon seasons. Rainfall amount (less than 10 mm and greater than 150 mm) and rainfall frequency was underestimated by CHIRPS in all basins, while the overestimated rainfall was between 10 and 100 mm in all basins except Kamala. Additionally, CHIRPS overestimated dry days and maximum consecutive dry days in the study area. Our study suggests that CHIRPS rainfall products cannot supplant the ground-based observations but complement rain-gauge networks. However, the reliability of this product in capturing local extreme events (such as floods and droughts) seems less prominent. A high-quality rain gauge network is essential to enhance the accuracy of satellite estimations.

As a new information processing science, neural network is an abstraction and simulation of some basic functions of the human brain. It is based on the working model of the human brain to study adaptive and non-process information processing technology. The uniqueness of this working mechanism is that the processing function of the body is reflected in the role of a large number of neurons in the network. Starting from modeling the structure of the human brain and the function of a single neuron, the information processing in the human brain can be modeled. In the climate during the historical heavy rain, study various relationships and factors, analyze its natural background and conditions to make it more severe, then estimate what will happen at a certain point in time, and place the scene when the location reaches its maximum. The time-division topographic improvement factor method can quantitatively divide the typhoon A and typhoon B rainstorm into two parts: the concentrated rain component and the field rain component, and the spatial distribution of these two components can be obtained. The mountain’s impact on the 24-h rainfall of typhoon A reached 45%, while the 4-h rainfall increase of Typhoon B was 24%. For further analysis, the 24-/4-h convergent rainfall pattern derived from this summary can be moved to a specific area to estimate the 24-/4-h PMP. Improving English proficiency is an inevitable process for students to learn and use English, and language expression writing is also essential for social communication. In order to perform well in English class and improve students’ writing ability, teachers need to use more self-help teaching methods and compile students’ learning materials for a long time to truly improve students’ writing ability. [ABSTRACT FROM AUTHOR]

An approach to developing a blended satellite-rainfall dataset over Australia that could be suitable for operational use is presented. In this study, Global Satellite Mapping of Precipitation (GSMaP) satellite precipitation estimates were blended with station-based rain gauge data over Australia, using operational station data that has not been harnessed by other blended products. A two-step method was utilized. First, GSMaP satellite precipitation estimates were adjusted using rain gauge data through multiplicative ratios that were gridded using ordinary kriging. This step resulted in reducing dry biases, especially over topography. The adjusted GSMaP data was then blended with the Australian Gridded Climate Dataset (AGCD) rainfall analysis, an operational station-based gridded rain gauge dataset, using an inverse error variance weighting method to further remove biases. A validation that was performed using a 20-year range (2001 to 2020) showed the proposed approach was successful; the resulting blended dataset displayed superior performance compared to other non-gauge-based datasets with respect to stations as well as displaying more realistic patterns of rainfall than the AGCD in areas with no rain gauges. The average mean absolute error (MAE) against station data was reduced from 0.89 to 0.31. The greatest bias reductions were obtained for extreme precipitation totals and over mountainous regions, provided sufficient rain gauge availability. The newly produced dataset supported the identification of a general positive bias in the AGCD over the north-west interior of Australia.

Rainfall is one of the most important elements to determine the climate. Therefore, it is important to estimate its value accurately. The main purpose of this study is the evaluation of the TRMM (Tropical Rain Measurement Mission) 3B42 rainfall estimates, an exponential model and conceptual cloud model in Lake Urmia Basin. Therefore, this study focuses on the comparison of these methods to identify and select the most appropriate model for rainfall estimation in Lake Urmia Basin. The comparison are performed during the period 2007 to 2011 and the hourly rainfall, temperature, barometric pressure and dew point temperature, the three-hourly rainfall rate of TRMM 3B42-V6 at 0.25° resolution and thermal infrared images (TIR) of Meteosat 7 at six-hour intervals are used. The results indicated acceptable match of estimated rainfall with rain-gauge data. Comparison of three methods of rainfall estimation shows that exponential model has the determination coefficient (equal to 0.61). In addition to the high correlation, due to low levels of RMSE and MAE (respectively 1.58 and 1.01), has a good performance to estimate rainfall in this basin. Therefore, this model can introduced as the most appropriate model for estimating rainfall in Lake Urmia basin.

In the framework of Soil Moisture and Ocean Salinity (SMOS) Calibration/Validation (Cal/Val) activities, this study addresses the use of the PERSIANN-CCS(1)database in hydrological applications to accurately simulate a whole SMOS pixel by representing the spatial and temporal heterogeneity of the soil moisture fields over a wide area (50x50 km(2)). The study focuses on the Valencia Anchor Station (VAS) experimental site, in Spain, which is one of the main SMOS Cal/Val sites in Europe. A faithful representation of the soil moisture distribution at SMOS pixel scale (50x50 km(2)) requires an accurate estimation of the amount and temporal/spatial distribution of precipitation. To quantify the gain of using the comprehensive PERSIANN database instead of sparsely distributed rain gauge measurements, comparisons between in situ observations and satellite rainfall data are done both at point and areal scale. An overestimation of the satellite rainfall amounts is observed in most of the cases (about 66%) but the precipitation occurrences are in general retrieved (about 67%). To simulate the high variability in space and time of surface soil moisture, a Soil Vegetation Atmosphere Transfer (SVAT) model - ISBA (Interactions between Soil Biosphere Atmosphere) is used. The interest of using satellite rainfall estimates as well as the influence that the precipitation events can induce on the modelling of the water content in the soil is depicted by a comparison between different soil moisture data. Point-like and spatialized simulated data using rain gauge observations or PERSIANN - CCS database as well as ground measurements are used. It is shown that a good adequacy is reached in most part of the year, the precipitation differences having less impact upon the simulated soil moisture. The behaviour of simulated surface soil moisture at SMOS scale is verified by the use of remote sensing data from the Advanced Microwave Scanning Radiometer on Earth observing System (AMSR-E). We show that the PERSIANN database provides useful information at temporal and spatial scales in the context of soil moisture retrieval.

Rainfall estimation based on the impact of rain on electromagnetic waves is a novel methodology that has had notable advancements during the last few years. Many studies conducted on this topic in the past considered only the electromagnetic waves with frequencies greater than 10 GHz since the rainfall impact on the electromagnetic wave attenuation is reduced at lower frequencies. Over the last few years, some authors have demonstrated that there can be a non-negligible attenuation even on the signals received on a global system for mobile communications mobile terminal in presence of rain. In this paper, we propose a new classification method based on a probabilistic neural network to obtain an accurate classification between four rainfall intensities (no rain, weak rain, moderate rain, and heavy rain). The innovative rainfall classification method is based on three received signal level (RSL) local features of the 4G/LTE: the instantaneous RSL, the average RSL value, and its variance calculated by using a sliding window. The proposed method exhibits good performance, obtaining an overall correct classification rate of 96.7%. Almost all papers on this topic present in the literature focus on electromagnetic waves with frequencies greater than 10 GHz, in which the rain impact is more relevant, according to the rain attenuation model. However, only the 4G/LTE signal has such widespread geographic coverage, so the proposed classification method can provide noticeable improvements in the creation of rainfall maps with higher spatial resolution.

The current paper introduces a new multilayer perceptron (MLP) and support vector machine (SVM) based approach to improve daily rainfall estimation from the Meteosat Second Generation (MSG) data. In this study, the precipitation is first detected and classified into convective and stratiform rain by two MLP models, and then four multi-class SVM algorithms were used for daily rainfall estimation. Relevant spectral and textural input features of the developed algorithms were derived from the spectral MSG SEVIRI radiometer channels. The models were trained using radar rainfall data set colected over north Algeria. Validation of the proposed daily rainfall estimation technique was performed by rain gauge network data set recorded over north Algeria. Thus, several statistical scores were calculated, such as correlation coefficient (r), root mean square error (RMSE), mean error (Bias), and mean absolute error (MAE). The findings given by: (r = 0.97, bias = 0.31 mm, RMSE = 2.20 mm and MAE = 1.07 mm), showed a quite satisfactory relationship between the estimation and the respective observed daily precipitation. Moreover, the comparison of the results with those of two advanced techniques based on random forests (RF) and weighted 'k' nearest neighbor (WkNN) showed higher accuracy obtained by the proposed model. [ABSTRACT FROM AUTHOR]