Model Precipitation Forecasts Research Articles

New method of precipitation forecast model and validation

There is a lot of time series data in many realistic sectors with different forecast techniques over the years. However there is no unanimous conclusion on forecast techniques such as individual forecasts Autoregressive, Moving averages, Autoregressive Moving average, Autoregressive Integrated Moving average, Artificial Neural Network, Long Short Term Memory network and Auto-Regressive Conditional Heteroscedasticity / Generalized Autoregressive Conditional Heteroskedasticity and combination of forecast (simple Average of forecasts, Minimum variance method, and Regression method of the combine). The most empirical hydrological time series models do not accurately forecast the weather. This paper focuses on a comparative study of different existing individual and combination forecasts with the proposed Hybrid Stochastic Model (HSM) forecast procedure. For this we consider a hydrological time series data of the Indian subcontinent to test the proposed forecast model. As a whole in comparison to all other traditional model's contributions accuracy, the proposed model performed well, and also we examined the model's dimension reduction approach to choose an optimum number of forecast techniques to be included in the model to yield the best forecasts.

A Novel Response Priority Framework for an Urban Coastal Catchment Using Global Weather Forecasts‐Based Improved Flood Risk Estimates

AbstractA novel flood risk forecasting‐based response priority framework is proposed in this study, which will facilitate efficient decision‐making ahead of an extreme precipitation event. This study is demonstrated over a highly flood‐prone and geomorphologically diverse coastal catchment in Mumbai city, the financial capital of India, and is subjected to high spatio‐temporal variability of precipitation. A regional precipitation forecast model is developed, where a quantile regression (QR)‐based statistical approach is applied to the global weather forecasts for improvement in finer resolution extreme precipitation estimates. A QR is performed between the observed synoptic‐scale predictors obtained from ERA‐Interim Reanalysis data set and the observed subdaily precipitation within the study area. Subsequently, a set of reliably well‐simulated forecasted predictors obtained from the ensemble members of the Global Ensemble Forecast System are used in the established relationship to obtain extreme precipitation forecasts for higher quantile levels. The forecasted precipitation data serves as an essential input to a comprehensive hydrodynamic flood modeling framework. The flood inundation maps corresponding to different quantiles of forecasted precipitation are quantified to identify the flood hotspots. The set of flood inundations maps developed for different quantiles are utilized to demarcate the areas based on critical response time, that is, the time required by each cell within the study area to be inundated to a certain threshold depth, and the response priority levels for each cell are determined. This information is combined with socioeconomic vulnerability (SEV), a critical component of flood risk, to derive conjugated SEV‐based response priority maps.

A precipitation forecast model with a neural network and improved GPT3 model for Japan

A precipitation forecast model with a neural network and improved GPT3 model for Japan

Seasonal streamflow forecasting in South America’s largest rivers

Study regionSouth America large basins (>5000 km²) Study focusThis work represents a first assessment of seasonal streamflow forecasts in South America based on a continental-scale application of a large scale hydrologic-hydrodynamic model and ECMWF's seasonal forecasting system precipitation forecasts (SEAS5-SSF) with bias correction. Seasonal streamflow forecasts were evaluated against a reference model run. Forecast skill was estimated relative to the Ensemble Streamflow Prediction (ESP) method. New hydrological insightsWe observed that bias correction was essential to obtain positive skill of SEAS5-SSF over ESP, which remained a hard to beat benchmark, especially in regions with high seasonality, and highly dependent on initial conditions. SEAS5-SSF skill was found to be dependent on initialization month, basin and lead time. Rivers where the skill is higher were Amazon, Araguaia, Tocantins and Paraná.

Evaluation of a novel hybrid lion swarm optimization – AdaBoostRegressor model for forecasting monthly precipitation

Precipitation is essential for crop production, water resource management, and other activities. Since precipitation has unpredictable sequential and seasonal characteristics, it is extremely difficult to forecast precipitation. Hence, a better forecasting model will lead to an early warning that minimizes the risk to life and property. Many approaches, such as statistical, Machine Learning (ML), Deep Learning (DL), and hybrid algorithms, are used to build precipitation forecasting systems. Feature selection is especially important for selecting past values (lags) and to eliminate irrelevant and/or redundant features. The optimization algorithm helps identify correct time series lags because it does not make data assumptions. In this work, a hybrid long-term (1-month ahead) precipitation forecast model (ABR-LSO) is proposed by integrating the AdaBoostRegressor (ABR) with the Lion Swarm Optimization (LSO) algorithm. Time series precipitation data for the period 1989–2019 for Idukki, Kerala, India. The ABR model considered 12 antecedent rainfall records as its input features and predicted the rainfall data for the next time step (t + 1), whereas the ABR-LSO hybrid model optimized the input features with LSO. The efficacy of the forecast models is evaluated using statistical indices like mean absolute error (MAE) and root mean square error (RMSE). It has been noted that in models 1, 2, and 3, the LSO-ABR model produced good MAE above ABR model improvements of 5.08 %, 4 %, and 6.12 %, respectively. Precipitation forecasts will help the administrators devise disaster prevention mechanisms against flooding in the downstream areas.

Forecasting the July Precipitation over the Middle-Lower Reaches of the Yangtze River with a Flexible Statistical Model

The multiple regression method is still an important tool for establishing precipitation forecast models with a lead time of one season. This study developed a flexible statistical forecast model for July precipitation over the middle-lower reaches of the Yangtze River (MLYR) based on the prophase winter sea surface temperature (SST). According to the characteristics of observed samples and related theoretical knowledge, some special treatments (i.e., more flexible and better-targeted methods) were introduced in the forecast model. These special treatments include a flexible MLYR domain definition, the extraction of indicative signals from the SST field, artificial samples, and the amplification of abnormal precipitation. Rolling forecast experiments show that the linear correlation between prediction and observation is around 0.5, more than half of the abnormal precipitation years can be successfully predicted, and there is no contradictory prediction of the abnormal years. These results indicate that the flexible statistical forecast model is valuable in real-life applications. Furthermore, sensitivity experiments show that forecast skills without these special treatments are obviously decreased. This suggests that forecast models can benefit from using statistical methods in a more flexible and better-targeted way.

Study on Short-term and Imminent Precipitation Prediction Model and Accuracy Evaluation Based on Depth Learning

In the field of weather forecasting, short-term precipitation forecast is still a global problem. The complicated mechanism of convective weather system, the medium and small scale weather system brings great difficulties to the forecast, which is difficult to detect, understand and forecast. The application of short-term precipitation forecast is still in the initial stage. At present, the optical flow method based on radar echo is widely used in short-term and imminent precipitation prediction, but the optical flow method has obvious limitations. In order to improve the forecast precision of short-term and imminent precipitation forecast model, this study introduces the method of depth learning to build the forecast model. The application of depth learning to short-term precipitation forecasting is still in its initial stage. We analyze the existing models may exist in the shortcomings of the proposed in-depth learning model combination of different ideas. By combining different network structures (such as CONVLSTM, CONVGRU, etc.), we can get a better combined forecasting model. It has certain application value.

Методики краткосрочного прогнозирования стока рек бассейна Камы

For the rivers of the Kama basin, two methods have been developed for daily short-term forecasting of average daily water discharges with a lead time of 1, 2, and 3 days. The first technique uses the HBV-96 river runoff formation conceptual model and the subsequent correction of forecasts outputs. The second technique is based on the hydrograph extrapolation method and is limited to the data of hydrological observations available by the date of the forecast. For the first technique, the expected values of the COSMO-Ru meteorological model temperature and precipitation forecasts are used for the lead-time period of the hydrological forecast. The influence of meteorological forecast errors on the accuracy of river streamflow forecasts was analyzed. The results of the verification based on an independent sample showed that, in general, both methods give a satisfactory forecast quality. The technique that uses the runoff formation model and meteorological information is slightly more accurate than the hydrograph extrapolation method, and its advantage increases with increasing forecast lead time. Keywords: Kama basin, water discharge, hydrological model, correction, extrapolation, verification, forecast quality

Summer Precipitation Forecast Using an Optimized Artificial Neural Network with a Genetic Algorithm for Yangtze-Huaihe River Basin, China

Owing to the complexity of the climate system and limitations of numerical dynamical models, machine learning based on big data has been used for climate forecasting in recent years. In this study, we attempted to use an artificial neural network (ANN) for summer precipitation forecasts in the Yangtze-Huaihe River Basin (YHRB), eastern China. The major ANN employed here is the standard backpropagation neural network (BPNN), which was modified for application to the YHRB. Using the analysis data of precipitation and the predictors/factors of atmospheric circulation and sea surface temperature, we calculated the correlation coefficients between precipitation and the factors. In addition, we sorted the top six factors for precipitation forecasts. In order to obtain accurate forecasts, month (factor)-to-month (precipitation) forecast models were applied over the training and validation periods (i.e., summer months over 1979–2011 and 2012–2019, respectively). We compared the standard BPNN with the BPNN using a genetic algorithm-based backpropagation (GABP), support vector machine (SVM) and multiple linear regression (MLR) for the summer precipitation forecast after the model training period, and found that the GABP method is the best among the above methods for precipitation forecasting, with a mean absolute percentage error (MAPE) of approximately 20% for the YHRB, which is substantially lower than the BPNN, SVM and MLR values. We then selected the best summer precipitation forecast of the GABP month-to-month models by summing up monthly precipitation, in order to obtain the summer scale forecast, which presents a very successful performance in terms of evaluation measures. For example, the basin-averaged MAPE and anomaly rate reach 4.7% and 88.3%, respectively, for the YHRB, which can be a good recommendation for future operational services. It appears that sea surface temperatures (SST) in some key areas dominate the factors for the forecasts. These results indicate the potential of applying GABP to summer precipitation forecasts in the YHRB.

Comparison of forecasting accuracy for Madden Julian Oscillation (MJO) and Convectively Coupled Equatorial Waves (CCEW) using Tropical Rainfall Measuring Mission (TRMM) and ERA-Interim precipitation forecast data for Indonesia

Tropical Rainfall Measuring Mission (TRMM) and ERA-Interim forecast data analyzed using second-order autoregressive AR(2) and space-time-spectra analysis methods (respectively) revealed contrasting results for predicting Madden Julian Oscillation (MJO) and Convectively Coupled Equatorial Waves (CCEW) phenomena over Indonesia. This research used the same 13-year series of daily TRMM 3B42 V7 derived datasets and ERA-Interim reanalysis model datasets from the European Center for Medium-Range Weather Forecasts (ECMWF) for precipitation forecasts. Three years (2016 to 2018) of the filtered 3B42 and ERA-Interim forecast data was then used to evaluate forecast accuracy by looking at correlation coefficients for forecast leads from day +1 through day +7. The results revealed that rainfall estimation data from 3B42 provides better results for the shorter forecast leads, particularly for MJO, equatorial Rossby (ER), mixed Rossby-gravity (MRG), and inertia-gravity phenomena in zonal wavenumber 1 (IG1), but gives poor correlation for Kelvin waves for all forecast leads. A consistent correlation for all waves was achieved from the filtered ERA-Interim precipitation forecast model, and although this was quite weak for the first forecast leads it did not reach a negative correlation in the later forecast leads except for IG1. Furthermore, Root Mean Square Error (RMSE) was also calculated to complement forecasting skills for both data sources, with the result that residual RMSE for the filtered ERA-Interim precipitation forecast was quite small during all forecast leads and for all wave types. These findings prove that the ERA-Interim precipitation forecast model remains an adequate precipitation model in the tropics for MJO and CCEW forecasting, specifically for Indonesia.

Application of an Improved Analog-Based Heavy Precipitation Forecast Model to the Yangtze—Huai River Valley and Its Performance in June–July 2020

Application of an Improved Analog-Based Heavy Precipitation Forecast Model to the Yangtze—Huai River Valley and Its Performance in June–July 2020

Assessing the role of short-term weather forecasts in fire manager tactical decision-making: a choice experiment

BackgroundWeather plays an integral role in fire management due to the direct and indirect effects it has on fire behavior. However, fire managers may not use all information available to them during the decision-making process, instead utilizing mental shortcuts that can bias decision-making. Thus, it is important to evaluate if (and how) fire managers use information like weather forecasts when making tactical decisions. We explore USDA Forest Service fire manager confidence in relative humidity, precipitation, and wind models. We then use a choice experiment where key weather attributes were varied to explore how sensitive fire managers were to changes in specific weather variables when choosing to directly or indirectly attack a fire that is transitioning to extended attack.ResultsRespondents were less confident in the accuracy of wind and precipitation forecasts than relative humidity or weather forecasts more generally. The influence of weather information on the decision depended on the framing used in the choice experiment; specifically, whether respondents were told the initial strategy had been to directly or indirectly attack the fire. Across conditions, fire managers generally preferred to indirectly attack the fire. Decisions about the tactics to apply going forward were more sensitive to time in season when the fire was occurring and wind and precipitation forecasts than to other attributes.ConclusionsThe results have implications for the design of decision support tools developed to support fire management. Results suggest how fire managers’ use of fire weather information to evaluate forecast conditions and adjust future management decisions may vary depending on the management decision already in place. If fire weather-based decision support tools are to support the use of the best available information to make fire management decisions, careful attention may be needed to debias any effect of prior decisions. For example, decision support tools may encourage users to “consider the opposite,” i.e., consider if they would react differently if different initial decision with similar conditions were in place. The results also highlight the potential importance of either improving wind and precipitation forecast models or improving confidence in existing models.

Copula based post-processing for improving the NMME precipitation forecasts

Using reliable and timely precipitation forecasts on a monthly or seasonal scale could be useful in many water resources management planning, especially in countries facing drought challenges. Amongst many, the North American Multi-Model Ensemble (NMME) is one of the most well-known models. In this study, a Bayesian method based on Copula functions has been applied to improve NMME precipitation forecasts. This method is based on the existence of a correlation between the raw forecast and observational data. Two main factors affect the results of rainfall improvement based on the selected method. This research has presented innovative methods in these regards namely; 1) the approach of selecting the appropriate statistical distribution for variables and 2) the selection method of improved data according to the conditional probability distribution functions (CPDF). To evaluate the effectiveness of the statistical distribution, firstly the precipitation forecast improvement model has been developed based on the application of parametric (Exponential, Normal, Gamma, LogNormal and General Exreteme Value (GEV)) and non-parametric distributions (Standard Normal Kernel). Then the novel mixed distribution function based on GEV parametric distribution and Standard Normal Kernel (non-parametric distribution) has been suggested. As the second aim, a new method for selecting improved data based on the center of mass of estimated CPDF is presented. The evaluation of the proposed method for estimating the statistical distribution of data and improving the forecast precipitation by the NMME model has been performed in Sistan and Baluchestan province in Iran. In this regard, the data of 1982–2010 for the calibration period and the data of 2012–2016 for the validation of the results have been used. According to the results, the non-parametric distribution best fitted with the data in the time series and selecting the appropriate bandwidth increased the efficiency of this distribution. Besides, due to the weakness of non-parametric distributions in the boundaries, the use of GEV distribution with a high ability to estimate boundary conditions as semi-parametric distribution, led to improved performance of the proposed distribution. Finally, the selection of the improved data based on the center of the mass method has efficiently provided much improvement compared to the maximum likelihood method commonly used.

A quantitative precipitation forecast model using convective-cloud tracking in satellite thermal infrared images and adaptive regression: a case study along East Coast of India

The current work addresses issues related to quantitative estimation of precipitation caused by convective clouds, using thermal infrared images, and adaptive regression modeling. The developed methodology has been implemented on the Indian sector during the period 22–25 October 2013. The importance of the developed methodology lies in the fact that the information obtained from it can facilitate further studies intended for the prediction of flood events. This study is the continuation of existing work of identification of convective clouds and the analysis of the Mesoscale Convective Systems (MCS). In the current work, forecast of rainfall in terms of millimeter has been proposed. The entire work has been carried out on thermal infrared (TIR) images obtained from geostationary satellites and the results have been validated by actual rainfall data measured by rain gauges. The results obtained from the developed methodology were found to be fairly close to actual values.

Uncertainty reduction in quantitative precipitation prediction by tuning of Kain–Fritch scheme input parameters in the WRF model using the simulated annealing optimization method

AbstractIn the study, using the simulated annealing (SA) optimization method, tried to reduce the uncertainty of the prediction of the Weather Research and Forecasting (WRF) numerical model in convective rainfall forecasts. To this end, three parameters Pd, Pe and Ph in Kain–Fritch convective scheme that's related to downdraft mass flux, entrainment mass flux and starting height of downdraft above updraft source layer, respectively, are optimized using SA algorithm to achieve better values. Two nested domain were used in the study with 30 and 10 km resolution which inner domain cover southern coasts of the Caspian Sea for the study area. Runtime of the model was 36 hr with the first 12 hr spin‐up time. Study period selected October 12, 2012 for training algorithm and October 8, 2015 for test run. After 100 iteration of the algorithm, 1, 1 and 50 was obtained for Pd, Pe and Ph respectively, while default values of these parameters was 0, 0, and 150. Results show that in both cases, model with default values underestimates rainfall amount and after optimization, model performance improves. Also, spatial distribution of the model precipitation forecast was less than observations and after optimization, spatial distribution improves. Statistical analysis of results indicated Mean Bias (MB) and root mean square error (RMSE) in training case were −4.6 and 17.1 in model with default values and became −3.7 and 14.3 in optimized one, respectively. Also, MB and RMSE for test case increased to −8.4 and 23.8, respectively, in model result with optimized parameters from −10.6 to 29.4 in model with default parameters.

Sugarcane decision-making support using Eta Model precipitation forecasts

Agricultural activity is largely influenced by climatic conditions. Rainfall is essential for crop production, and precipitation events also interfere with soil preparation, planting, application of pesticides and harvesting. Weather forecast models are tools to facilitate decision making for agricultural activities, hence high accuracy is desired. Farmers often criticize the accuracy of weather forecasts, which sometimes fail to predict precipitation events, leading to yield loss and environmental harm. In this study, precipitation forecasts of the Eta Model were evaluated for 28 of Brazil’s most productive sugarcane areas, considering a grid of 15 × 15 km. Using a combination of different indicators of forecast success, observed and forecasted daily precipitation data were compared for consecutive days of all 10-day periods in a course of 6 years (2005–2010). Skill scores and performance diagrams based on the indicators were used to evaluate the goodness and robustness of the model forecasts. The Eta Model forecasts showed overall accuracies ranging between 55 and 71% for the Atlantic forest biomes (located North-West and South-East of Sao Paulo) and the Cerrado biomes (located in the Goias State and in the Center-North Sao Paulo State), respectively. The forecasts were most reliable for up to 4 days, showing an accuracy of 60%. Forecasts for periods of more than 4 days had an average accuracy of 40–50%. The probability of detecting rainfall correctly was the strongest characteristic of Eta Model, with more than 70% hits.

The potential for uncertainty in Numerical Weather Prediction model verification when using solid precipitation observations

AbstractPrecipitation forecasts made by Numerical Weather Prediction (NWP) models are typically verified using precipitation gauge observations that are often prone to the wind‐induced undercatch of solid precipitation. Therefore, apparent model biases in solid precipitation forecasts may be due in part to the measurements and not the model. To reduce solid precipitation measurement biases, adjustments in the form of transfer functions were derived within the framework of the World Meteorological Organization Solid Precipitation Inter‐Comparison Experiment (WMO‐SPICE). These transfer functions were applied to single‐Alter shielded gauge measurements at selected SPICE sites during two winter seasons (2015–2016 and 2016–2017). Along with measurements from the WMO automated field reference configuration at each of these SPICE sites, the adjusted and unadjusted gauge observations were used to analyze the bias in a Global NWP model precipitation forecast. The verification of NWP winter precipitation using operational gauges may be subject to verification uncertainty, the magnitude and sign of which varies with the gauge‐shield configuration and the relation between model and site‐specific local climatologies. The application of a transfer function to alter‐shielded gauge measurements increases the amount of solid precipitation reported by the gauge and therefore reduces the NWP precipitation bias at sites where the model tends to overestimate precipitation, and increases the bias at sites where the model underestimates the precipitation. This complicates model verification when only operational (non‐reference) gauge observations are available. Modelers, forecasters, and climatologists must consider this when comparing modeled and observed precipitation.

Improving sub-seasonal forecast skill of meteorological drought: a weather pattern approach

Abstract. Dynamical model skill in forecasting extratropical precipitation is limited beyond the medium-range (around 15 d), but such models are often more skilful at predicting atmospheric variables. We explore the potential benefits of using weather pattern (WP) predictions as an intermediary step in forecasting UK precipitation and meteorological drought on sub-seasonal timescales. Mean sea-level pressure forecasts from the European Centre for Medium-Range Weather Forecasts ensemble prediction system (ECMWF-EPS) are post-processed into probabilistic WP predictions. Then we derive precipitation estimates and dichotomous drought event probabilities by sampling from the conditional distributions of precipitation given the WPs. We compare this model to the direct precipitation and drought forecasts from the ECMWF-EPS and to a baseline Markov chain WP method. A perfect-prognosis model is also tested to illustrate the potential of WPs in forecasting. Using a range of skill diagnostics, we find that the Markov model is the least skilful, while the dynamical WP model and direct precipitation forecasts have similar accuracy independent of lead time and season. However, drought forecasts are more reliable for the dynamical WP model. Forecast skill scores are generally modest (rarely above 0.4), although those for the perfect-prognosis model highlight the potential predictability of precipitation and drought using WPs, with certain situations yielding skill scores of almost 0.8 and drought event hit and false alarm rates of 70 % and 30 %, respectively.

МЕЗОМАСШТАБНАЯ МЕТЕОРОЛОГИЧЕСКАЯ МОДЕЛЬ TSUNM3 ДЛЯ ИССЛЕДОВАНИЯ И ПРОГНОЗИРОВАНИЯ СОСТОЯНИЯ МЕТЕОПАРАМЕТРОВ ПРИЗЕМНОГО СЛОЯ АТМОСФЕРЫ НАД КРУПНЫМ НАСЕЛЕННЫМ ПУНКТОМ

The paper describes the mathematical formulation and numerical method of the TSUNM3 high-resolution mesoscale meteorological model being developed at Tomsk State University. The model is nonhydrostatic and includes three-dimensional nonstationary equations of hydrothermodynamics of the atmospheric boundary layer with parameterization of turbulence, moisture microphysics, long-wave and short-wave (solar) radiation, and advective and latent heat flows in the atmosphere and at the boundary of its interaction with the underlying surface. The numerical algorithm is constructed using structured grids with uniform spacing in horizontal directions and condensing to the Earth surface in the vertical direction. When approximating the differential formulation of the problem, the finite volume method with the second order approximation in the spatial variables is used. Explicit-implicit approximations in time (Adams–Bashforth and Crank–Nicolson) are used to achieve second-order accuracy in time. The paper presents results of numerical forecasting of the main meteorological parameters of the atmosphere (temperature, humidity, wind speed and direction) and precipitation in different seasons in the Siberian region. The models were tested with the help of observations obtained using the Volna-4M sodar, MTR-5 temperature profile meter, and Meteo-2 ultrasonic weather stations of the Atmosfera Collective Use Center. The improved TSUNM3 model is shown to adequately reflect the precipitation time and intensity. However, in some cases, the times of its beginning and end do not always coincide, the difference can reach several hours. The precipitation phase state is reflected reliably. Over 70% of precipitation cases are confirmed by numerical calculations. The model satisfactorily predicts temperature and humidity characteristics. The quality of the precipitation forecast model is comparable to the modern mesoscale models, such as the Weather Research and Forecasting (WRF) model.

Algorithm to Predict the Rainfall Starting Point as a Function of Atmospheric Pressure, Humidity, and Dewpoint

Forecasting extreme precipitations is one of the main priorities of hydrology in Latin America and the Caribbean (LAC). Flood damage in urban areas increases every year, and is mainly caused by convective precipitations and hurricanes. In addition, hydrometeorological monitoring is limited in most countries in this region. Therefore, one of the primary challenges in the LAC region the development of a good rainfall forecasting model that can be used in an early warning system (EWS) or a flood early warning system (FEWS). The aim of this study was to provide an effective forecast of short-term rainfall using a set of climatic variables, based on the Clausius–Clapeyron relationship and taking into account that atmospheric water vapor is one of the variables that determine most meteorological phenomena, particularly regarding precipitation. As a consequence, a simple precipitation forecast model was proposed from data monitored at every minute, such as humidity, surface temperature, atmospheric pressure, and dewpoint. With access to a historical database of 1237 storms, the proposed model allows use of the right combination of these variables to make an accurate forecast of the time of storm onset. The results indicate that the proposed methodology was capable of predicting precipitation onset as a function of the atmospheric pressure, humidity, and dewpoint. The synoptic forecast model was implemented as a hydroinformatics tool in the Extreme Precipitation Monitoring Network of the city of Queretaro, Mexico (RedCIAQ). The improved forecasts provided by the proposed methodology are expected to be useful to support disaster warning systems all over Mexico, mainly during hurricanes and flashfloods.