Daily Rainfall Time Series Research Articles

An ensemble neural network approach for space–time landslide predictive modelling

There is an urgent need for accurate and effective Landslide Early Warning Systems (LEWS). Most LEWS are currently based on temporally-aggregated measures of rainfall derived from either in-situ measurements or satellite-based rainfall estimates. Relying on a summary metric of precipitation may not capture the complexity of the rainfall signal and its dynamics in space and time in triggering landslides. Here, we present a proof-of-concept for constructing a LEWS based on an integrated spatio-temporal modelling framework. Our proposed methodology builds upon a recent approach that uses a daily rainfall time series instead of the traditional scalar aggregation. Specifically, we partition the study area into slope units and use a Gated Recurrent Unit (GRU) to process satellite-derived rainfall time series and combine the output features with a second neural network (NN) tasked with capturing the effect of terrain characteristics. To assess if our approach enhances accuracy, we applied it in Vietnam and benchmarked it against a modelling counterpart where we replaced the rainfall time series with the corresponding scalar representative of the cumulated precipitation. The corresponding duration was set at 14 days as it proved to produce the best performance. Our results show that our protocol leads to better performance in hindcasting landslides when making use of the rainfall as a continuous signal over time. While not tested here, our approach can be extended to rainfall obtained from weather forecasts, potentially leading to actual landslide forecasts.

Nonlinear hydrological time series modeling to forecast river level dynamics in the Rio Negro Uruguay basin.

Floods significantly impact the well-being and development of communities. Hence, understanding their causes and establishing methodologies for risk prevention is a critical challenge for effective warning systems. Complex systems such as hydrological basins are modeled through hydrological models that have been utilized to understand water recharge of aquifers, available volume of dams, and floods in diverse regions. Acquiring real-time hydrometeorological data from basins and rivers is vital for establishing data-driven-based models as tools for the prediction of river-level dynamics and for understanding its nonlinear behavior. This paper introduces a hydrological model based on a multilayer perceptron neural network as a useful tool for time series modeling and forecasting river levels in three stations of the Rio Negro basin in Uruguay. Daily time series of river levels and rainfall serve as the input data for the model. The assessment of the models is based on metrics such as the Nash-Sutcliffe coefficient, the root mean square error, percent bias, and volumetric efficiency. The outputs exhibit varying model performance and accuracy during the prediction period across different sub-basin scales, revealing the neural network's ability to learn river dynamics. Lagged time series analysis demonstrates the potential for chaos in river-level time series over extended time periods, mainly when predicting dam-related scenarios, which shows physical connections between the dynamical system and the data-based model such as the evolution of the system over time.

Evaluating Rainwater Harvesting Systems for Water Scarcity Mitigation in Small Greek Islands under Climate Change

Rainfall variability, exacerbated by climate change, poses significant challenges to water resource management, particularly in regions prone to intense droughts and floods. The Greek islands, characterized by poor water potential, face interannual water supply issues dating back to their earliest habitation. Rainwater harvesting (RWH) systems emerge as a promising solution to address water scarcity in these regions. This study simulates RWH systems for two small Greek islands, Fourni and Nisyros, representing similar rainfall regimes. Multiple scenarios are explored, and system reliability is assessed in light of simulated daily rainfall time series incorporating climate change projections. Utilizing eight low/medium (RCP 4.5) and eight high (RCP 8.5) emission scenarios over a future 35-year period, the study evaluates system reliability based on model parameters (collection area: 40 to 140 m2, rainwater tank volume: 5 to 30 m3, number of household members: 2, 3), with 30% coverage of total daily water demand (180 L/d). Negligible evapotranspiration effects are assumed due to closed-type tanks. Results indicate that the RWH system demonstrates high efficiency in general. The investigation for the future period revealed that the system’s performance varies, with instances where daily demand targets are not met, even with a 30 m3 tank. This research underscores the potential of RWH systems as a cost-effective “green” solution, particularly in regions with deficient rainfall regimes. It highlights the importance of localized water management strategies, reducing reliance on mainland water transportation, and assisting desalination unit operations. In conclusion, this study contributes to the assessment of RWH systems, demonstrating their viability as a sustainable water management solution in regions facing water scarcity, contingent on local rainfall conditions and system design parameters.

Modeling and Simulating Rainfall and Temperature Using Rotated Bivariate Copulas

Climate change is a significant environmental challenge that affects water resources, agriculture, health, and other aspects of human life. Bivariate modeling is a statistical method used to analyze the relationship between variables such as rainfall and temperature. The Pearson correlation coefficient, Kendall’s tau, or Spearman’s rank correlation are some measures used for bivariate modeling. However, copula functions can describe the dependence structure between two or more variables and can be effectively used to describe the relationship between rainfall and temperature. Despite the literature on bivariate modeling of rainfalls and temperature being extensive, finding flexible and sophisticated bivariate models is sometimes difficult. In this paper, we use rotated copula functions that can arrange any type of dependence that is empirically detected, especially negative dependence. The methodology is applied to an Italian municipality’s bivariate daily time series of rainfall and temperature. The estimated rotated copula is significant and, therefore, can be used for simulating the effects of extreme events.

Geo-statistical analysis of climatic variability and trend detection in the Hindu Kush region, North Pakistan.

This paper is an effort of geo-statistical analysis of rainfall variability and trend detection in the eastern Hindu Kush region located in the north-west of Pakistan. The eastern section of the HK region lies in the western part of Pakistan. Exploring rainfall variability and quantifying its trend and magnitude is one of the key indicators among all climatic parameters. In the study area, Pakistan Meteorology Department (PMD) has established seven meteorological stations: Drosh, Chitral, Dir, Timergara, Saidu Sharif, Malam Jabba, and Kalam. Daily, mean monthly, and mean annual rainfall time series data for all the met stations were geo-statistically analyzed in the GIS environment for detecting monthly and annual variability in rainfall, variability, and trend detection. Mann-Kendall (MK) and Theil-Sen's slope (TSS) statistical tests were applied to rainfall data. Initially, the MK test was applied for detection of trends and TSS test was used to quantify the change in magnitude. The results indicate that the rainfall variability in intensity and trend pattern detection. The analysis confirms that an extremely significant rainfall trend in the case of mean annual rainfall was predicted at Dir and Malam Jabba meteorological stations. Opposite to this, at Kalam and Chitral stations, a less significant rainfall trend was noted. In a similar context, no prominent rainfall trend has been found at Drosh, Timergara, and Saidu Sharif meteorological stations. Likewise, using TSS, an extremely negative variation in the magnitude of rainfall was verified at Kalam and Malam Jabba. However, a noteworthy positive change in rainfall magnitude has been noted at Dir and Saidu Sharif meteorological stations. The findings of this research have the potential to assist the decision and policy makers and academicians to think truly and conduct more scientific research studies to mitigate climate change.

Trends and impacts of climate change on crop production in Burkina Faso

Abstract Understanding past climate trends and their impacts in the Sahel region is fundamental for climate change (CC) adaptation and mitigation. This study analyses climate trends from 1961 to 2020 in three climatic zones in Burkina Faso and the impacts of CC on five major crops production. Long time series of daily rainfall and temperature data from National Meteorology Agency for the period 1961 to 2020 has been compiled. Crop production data (1984–2020) were retrieved from the agriculture department. Climate temporal variations in each climatic zone were analyzed using extreme climate indices and principal component analysis. Linear regression was used to assess climate impacts on crop production. The results showed a high rainfall variability and changes in temperature extremes in the three zones. The climate window, 1991–2020, was hotter than 1961–1990, while the last decade (2011–2020) was the wettest. Most climate indices (67%) showed significant correlations with crop yields. Dry spells, cool days, cold nights, average daily wet days and rainfall intensity showed positive and negative effects on maize, cowpea, millet and sorghum yields. This study highlights the importance of climate-smart policy promoting drought-resistant and short-duration varieties in addressing the adverse effects of CC on crop production.

Evaluating uncertainty in climate change impacts on peak discharge and flood volume in the Qaran Talar watershed, Iran

Abstract This study evaluates the effects of climate change (CC) on runoff properties over the case study of the Qaran Talar watershed located in Iran. To consider the two main sources of uncertainty, i.e., Greenhouse Gases emission (GHG) scenarios and outputs of Atmosphere-Ocean General Circulation Models (AOGCMs), a daily rainfall time series was generated for two future periods (2021-2050 and 2070-2099) at three risk levels of 0.1, 0.25 and 0.50. 22 AOGCMs outputs following two emission scenarios (RCP4.5 and RCP8.5) were used. The results showed that the uncertainty of climate change scenarios was primarily owing to the uncertainty of GCMs outputs. Regarding the 2021-2050 period, under both emission scenarios, the increases in peak discharge and flood volume (FV) were estimated to reach 70, 50, and 30% at three risk levels of 0.1, 0.25 and 0.50, respectively, compared to the recent past period. As the current century draws to a close, the difference between the results of the two emission scenarios becomes apparent so that in the far-future period (2070-2099), the RCP8.5 scenario estimates and FV more than the RCP4.5 scenario does. Additionally, it was found that the uncertainty caused by AOGCMs was more than that by GHG emission scenarios.

Analysis of the 160 years' time series of daily rainfall in Brisbane

Analysis of the 160 years' time series of daily rainfall in Brisbane

Analysis of the 160 years' time series of daily rainfall in Brisbane

Analysis of the 160 years' time series of daily rainfall in Brisbane

The Regional Diagnosis of Onset and Demise of the Rainy Season over Tropical and Subtropical Australia

Abstract In this paper, we introduce a novel strategy to robustly diagnose the onset and demise of the rainy season using daily observed rainfall over seven specific regions across Australia, as demarcated by the Natural Resource Management (NRM) agency of Australia. The methodology lies in developing an ensemble spread of the diagnosed onset and demise from randomly perturbing the observed daily time series of rainfall at synoptic scales to obtain a measure of the uncertainty of the diagnosis. Our results indicate that the spread of the ensemble in the diagnosis of the onset and demise dates of the rainy season is higher in the subtropical region than the tropical region. Secular change of earlier onset, later demise, longer length, and wetter season are also identified in many of these regions. The influence of the PDO at decadal scale, ENSO and Indian Ocean dipole at interannual scale, and MJO at intraseasonal scale also reveals significant influence on the evolution of the rainy season over these regions in Australia. Most important, the covariability of the onset date with the length of the season and seasonal rainfall anomaly of the season is highlighted as a valuable relationship that can be exploited for real-time monitoring and providing an outlook of the forthcoming rainy season, which could serve some of the NRM regions. Significance Statement We document the rainfall variability during the rainy season over tropical, subtropical, and semiarid regions of Australia and relate them to modes of climate variability spanning from intraseasonal to secular time scales. The study highlights the varied influence of the modes of climate variability on various aspects of the evolution of the rainy season, such as its onset and demise dates and the seasonal rainfall anomaly over these Australian regions. The study uses 114 years of data and shows that the variations in the length of the rainy season and its seasonal rainfall anomaly are strongly dictated by variations of rainy-season onset date. This provides a quick seasonal outlook of the forthcoming rainy season by just monitoring the onset-date evolution in these regions.

Daily streamflow forecasting by machine learning in Tra Khuc river in Vietnam

Precise streamflow prediction is crucial in the optimization of the distribution of water resources. This study develops the machine learning models by integrating recurrent gate unit (GRU) with bacterial foraging optimization (BFO), gray wolf optimizer (GWO), and human group optimization (HGO) to forecast the streamflow in the Tra Khuc River, Vietnam. For this purpose, the time series of daily rainfall and river flow at Son Giang station from 2000 to 2020 were employed to forecast the streamflow. The statistical indices, namely the root mean square error, the mean absolute error, and the coefficient of determination (R²), was utilized to evaluate the performance of the proposed models. The results showed that the three optimization algorithms (HGO, GWO, and BFO) effectively enhanced the performance of the GRU model.
 Moreover, among the four models (GRU, GRU-HGO, GRU-GWO, and GRU-BFO), the GRU-GWO model outperformed the other models with R² = 0.883. GRU-HGO achieved R² = 0.879, and GRU-BFO achieved R²=0.878. The results of this study showed that GRU combined with optimization algorithms is a reliable modeling approach in short-term flow forecasting.

Time-series analysis and statistical forecasting of daily rainfall in Kupang, East Nusa Tenggara, Indonesia

This pilot study presents a novel statistical time-series approach for analyzing daily rainfall data in Kupang, East Nusa Tenggara, Indonesia. By using the piecewise cubic hermite interpolation algorithm, we succeeded in filling in the null values in the daily rainfall time series. We then analyzed the monthly average and its pattern using the continuous wavelet transform (CWT) algorithm, which shows the strong annual pattern of rainfall in this region. In addition, we use the rainfall anomaly index (RAI) function to standardize daily rainfall as an indicator of dry/wet conditions in this region. Then we also use the daily RAI time-series objects from 1978 to 2020 for modeling and predicting daily RAI over the next year. The result is the root mean squared error (RMSE) of 0.8424041040593219. This Prophet model is also able to capture the linear trend of increasing drought throughout the study time period and the annual pattern of wet/dry conditions which is in accordance with previous study by Aldrian and Susanto (2003).

Simulating sediment discharge at water treatment plants under different land use scenarios using cascade modelling with an expert-based erosion-runoff model and a deep neural network

Abstract. Excessive sediment discharge in karstic regions can be highly disruptive to water treatment plants. It is essential for catchment stakeholders and drinking water suppliers to limit the impact of high sediment loads on potable water supply, but their strategic choices must be based on simulations integrating surface and groundwater transfers and taking into account possible changes in land use. Karstic environments are particularly challenging as they face a lack of accurate physical descriptions for the modelling process, and they can be particularly complex to predict due to the non-linearity of the processes generating sediment discharge. The aim of the study was to assess the sediment discharge variability at a water treatment plant according to multiple realistic land use scenarios. To reach that goal, we developed a new cascade modelling approach with an erosion-runoff geographic information system (GIS) model (WaterSed) and a deep neural network. The model was used in the Radicatel hydrogeological catchment (106 km2 in Normandy, France), where karstic spring water is extracted to a water treatment plant. The sediment discharge was simulated for five design storms under current land use and compared to four land use scenarios (baseline, ploughing up of grassland, eco-engineering, best farming practices, and coupling of eco-engineering/best farming practices). Daily rainfall time series and WaterSed modelling outputs extracted at connected sinkholes (positive dye tracing) were used as input data for the deep neural network model. The model structure was found by a classical trial-and-error procedure, and the model was trained on 2 significant hydrologic years. Evaluation on a test set showed a good performance of the model (NSE = 0.82), and the application of a monthly backward-chaining nested cross-validation revealed that the model is able to generalize on new datasets. Simulations made for the four land use scenarios suggested that ploughing up 33 % of grasslands would increase sediment discharge at the water treatment plant by 5 % on average. By contrast, eco-engineering and best farming practices will significantly reduce sediment discharge at the water treatment plant (respectively in the ranges of 10 %–44 % and 24 %–61 %). The coupling of these two strategies is the most efficient since it affects the hydro-sedimentary production and transfer processes (decreasing sediment discharge from 40 % to 80 %). The cascade modelling approach developed in this study offers interesting opportunities for sediment discharge prediction at karstic springs or water treatment plants under multiple land use scenarios. It also provides robust decision-making tools for land use planning and drinking water suppliers.

Impact of future climate change on malaria in West Africa

Understanding the regional impact of future climate change is one of the major global challenges of this century. This study investigated possible effects of climate change on malaria in West Africa in the near future (2006–2035) and the far future (2036–2065) under two representative concentration pathway (RCP) scenarios (RCP4.5 and RCP8.5), compared to an observed evaluation period (1981–2010). Projected rainfall and temperature were obtained from the coordinated regional downscaling experiment (CORDEX) simulations of the Rossby Centre Regional Atmospheric regional climate model (RCA4). The malaria model used is the Liverpool malaria model (LMM), a dynamical malaria model driven by daily time series of rainfall and temperature obtained from the CORDEX data. Our results highlight the unimodal shape of the malaria prevalence distribution, and the seasonal malaria transmission contrast is closely linked to the latitudinal variation of the rainfall. Projections showed that the mean annual malaria prevalence would decrease in both climatological periods under both RCPs but with a larger magnitude of decreasing under the RCP8.5. We found that the mean malaria prevalence for the reference period is greater than the projected prevalence for 6 of the 8 downscaled GCMs. The study enhances understanding of how malaria is impacted under RCP4.5 and RCP8.5 emission scenarios. These results indicate that the southern area of West Africa is at most risk of epidemics, and the malaria control programs need extra effort and help to make the best use of available resources by stakeholders.

Evaluation of Seven Gap-Filling Techniques for Daily Station-Based Rainfall Datasets in South Ethiopia

Meteorological stations, mainly located in developing countries, have gigantic missing values in the climate dataset (rainfall and temperature). Ignoring the missing values from analyses has been used as a technique to manage it. However, it leads to partial and biased results in data analyses. Instead, filling the data gaps using the reference datasets is a better and widely used approach. Thus, this study was initiated to evaluate the seven gap-filling techniques in daily rainfall datasets in five meteorological stations of Wolaita Zone and the surroundings in South Ethiopia. The considered gap-filling techniques in this study were simple arithmetic means (SAM), normal ratio method (NRM), correlation coefficient weighing (CCW), inverse distance weighting (IDW), multiple linear regression (MLR), empirical quantile mapping (EQM), and empirical quantile mapping plus (EQM+). The techniques were preferred because of their computational simplicity and appreciable accuracies. Their performance was evaluated against mean absolute error (MAE), root mean square error (RMSE), skill scores (SS), and Pearson’s correlation coefficients (R). The results indicated that MLR outperformed other techniques in all of the five meteorological stations. It showed the lowest RMSE and the highest SS and R in all stations. Four techniques (SAM, NRM, CCW, and IDW) showed similar performance and were second-ranked in all of the stations with little exceptions in time series. EQM+ improved (not substantial) the performance levels of gap-filling techniques in some stations. In general, MLR is suggested to fill in the missing values of the daily rainfall time series. However, the second-ranked techniques could also be used depending on the required time series (period) of each station. The techniques have better performance in stations located in higher altitudes. The authors expect a substantial contribution of this paper to the achievement of sustainable development goal thirteen (climate action) through the provision of gap-filling techniques with better accuracy.

A Nonparametric Procedure to Assess the Accuracy of the Normality Assumption for Annual Rainfall Totals, Based on the Marginal Statistics of Daily Rainfall: An Application to the NOAA/NCDC Rainfall Database

AbstractWe develop a nonparametric procedure to assess the accuracy of the normality assumption for annual rainfall totals (ART), based on the marginal statistics of daily rainfall. The procedure is addressed to practitioners and hydrologists that operate in data-poor regions. To do so we use 1) goodness-of-fit metrics to conclude on the approximate convergence of the empirical distribution of annual rainfall totals to a normal shape and classify 3007 daily rainfall time series from the NOAA/NCDC Global Historical Climatology Network database, with at least 30 years of recordings, into Gaussian (G) and non-Gaussian (NG) groups; 2) logistic regression analysis to identify the statistics of daily rainfall that are most descriptive of the G/NG classification; and 3) a random-search algorithm to conclude on a set of constraints that allows classification of ART samples on the basis of the marginal statistics of daily rain rates. The analysis shows that the Anderson–Darling (AD) test statistic is the most conservative one in determining approximate Gaussianity of ART samples (followed by Cramer–Von Mises and Lilliefors’s version of Kolmogorov–Smirnov) and that daily rainfall time series with fraction of wet days fwd < 0.1 and daily skewness coefficient of positive rain rates skwd > 5.92 deviate significantly from the normal shape. In addition, we find that continental climate (type D) exhibits the highest fraction of Gaussian distributed ART samples (i.e., 74.45%; AD test at α = 5% significance level), followed by warm temperate (type C; 72.80%), equatorial (type A; 68.83%), polar (type E; 62.96%), and arid (type B; 60.29%) climates.

Geostatistical and deterministic methods for rainfall interpolation in the Zayandeh Rud basin, Iran

ABSTRACT A new approach to select a model for fitting to a seasonal daily average rainfall semi-variogram was developed in this study by interpolating rainfall using geostatistical methods. Using two deterministic algorithms (Thiessen polygon, THI and inverse distance weighting, IDW) and two geostatistical algorithms (ordinary kriging, ORK and universal kriging, UNK) based on a rain gauge network, a 20-year daily and annual rainfall time series was generated for the Zayandeh Rud Dam Basin (Isfahan, Iran). Seven models were fitted to the experimental semi-variogram. Evaluation criteria – root mean square error (RMSE) and correlation coefficient (R) – identified that the Gaussian model is the best fit to the experimental semi-variogram. The ORK and UNK algorithms performed better in generating data when network sizes of 27, 15 or 10 gauges were used, whereas THI produced better results at a network size of five gauges. The results show that increasing the number of gauges does not necessarily produce better estimates, and stochastic methods are more sensitive than deterministic algorithms to network size.

Spatial analysis, assessment and mapping of flood hazard in the alluvial plains of Boumerzoug and Rhumel (city of Constantine, north-eastern Algeria): application to development and urban planning projects

This paper proposes an operational methodology for flood hazard assessment and mapping based on the use of two complementary methods: a geomorphologic analysis and a hydrological modelling method under GIS. The methodology has been applied to the urban and peri-urban areas of Constantine, extended over the flood plains of the Boumerzoug and Rhumel river catchment areas highly exposed to frequent flooding. The spatial analysis of the hazard factors revealed the natural predisposition of the two rivers plains of Rhumel and Boumerzoug to flooding. The hydrological modelling of flood phenomena by adjusting the time series of maximum daily rainfall and flow data of the five stations of the catchment areas to log-normal and normal laws allowed to obtain values of maximum daily rainfall and flows for different return periods of 5, 10, 50, 100 and 200 years. The obtained values are close to the recorded ones during the historical floods, which show the reliability of the used adjustment models. The obtained flood hazard map obtained on the basis of the identification of geomorphological units allowed to define the flood extension areas, the main overflow points and the degrees of hazard. This map can be useful for risk for decision about risk reduction and land use planning in the urban area of Constantine.

Characterization of the hydrodynamic functioning of the Degracias-Sic\xf3 Karst Aquifer, Portugal

A study on the complex processes of hydrogeological transformation of recharge into discharge within a karst system is described. Given the absence of boreholes or points of direct observation, the applied methodology is based on study of the natural response of the main outlet of the karst aquifer, considered here as a black-box. Time series analysis and recession curve analysis are applied to daily time series of rainfall as input, and spring discharge as the output of the aquifer. Daily variations of temperature and electrical conductivity of spring water are considered as well. The present study, an integrated research endeavor, is the first quantitative approach toward understanding the hydrogeological functioning of a large and not-well-known karst aquifer and improving existing knowledge about its internal physical characteristics. The case study is the Degracias-Sicó karst aquifer, in Portugal’s central-western region, which is relatively unknown from the hydrogeological perspective. The response of the karst aquifer’s main spring shows an important component of quick flow after heavy rainfall events and then a very slow recession. This bimodal behavior is proof of the significant spatial complexity and heterogeneity of the karst aquifer, marked by the presence of a conduit-dominated flow throughout a connected conduit network system and the dominance of base flow that confirms the aquifer’s large storage capacity. The results also demonstrate a significant component of delayed flow.

Characterizing the Uncertainty and Assessing the Value of Gap-Filled Daily Rainfall Data in Hawaii

AbstractAlmost all daily rainfall time series contain gaps in the instrumental record. Various methods can be used to fill in missing data using observations at neighboring sites (predictor stations). In this study, five computationally simple gap-filling approaches—normal ratio (NR), linear regression (LR), inverse distance weighting (ID), quantile mapping (QM), and single best estimator (BE)—are evaluated to 1) determine the optimal method for gap filling daily rainfall in Hawaii, 2) quantify the error associated with filling gaps of various size, and 3) determine the value of gap filling prior to spatial interpolation. Results show that the correlation between a target station and a predictor station is more important than proximity of the stations in determining the quality of a rainfall prediction. In addition, the inclusion of rain/no-rain correction on the basis of either correlation between stations or proximity between stations significantly reduces the amount of spurious rainfall added to a filled dataset. For large gaps, relative median errors ranged from 12.5% to 16.5% and no statistical differences were identified between methods. For submonthly gaps, the NR method consistently produced the lowest mean error for 1- (2.1%), 15- (16.6%), and 30-day (27.4%) gaps when the difference between filled and observed monthly totals was considered. Results indicate that gap filling prior to spatial interpolation improves the overall quality of the gridded estimates, because higher correlations and lower performance errors were found when 20% of the daily dataset is filled as opposed to leaving these data unfilled prior to spatial interpolation.