Analysis of Precipitation Variability using Memory Based Artificial Neural Networks

This article analyzes the variability in precipitation of the Barak river basin using memory-based ANN models called Gamma Memory Neural Network(GMNN) and genetically optimized GMNN called GMNN-GA for precipitation downscaling precipitation. GMNN having adaptive memory depth is capable techniques in modeling time varying inputs with unknown input characteristics, while an integration of the model with GA can further improve its performances. NCEP reanalysis and HadCM3A2 (a) scenario data are used for downscaling and forecasting precipitation series for Barak river basin. Model performances are analyzed by using statistical criteria, RMSE and mean error and are compared with the standard SDSM model. Results obtained by using 24 years of daily data sets show that GMNN-GA is efficient in downscaling daily precipitation series with maximum daily annual mean error of 6.78%. The outcomes of the study demonstrate that execution of the GMNN-GA model is superior to the GMNN and similar with that of the standard SDSM.

Evaluation of Generalized Extreme (GEV), Log-Pearson Type 3 (LP3), Pearson Type 3(P3) and Gumbel (EV1) Distributions for Development of IDF Equations for Warri, Nigeria

The application of Gumbel (EVI) to the development of rainfall intensity– duration – frequency (IDF) curves has often been criticized on theoretical and empirical grounds as it may underestimate the largest extreme rainfall amounts. The consequences of underestimation are economic losses, property damages, and loss of life. Therefore, it is important that water resources engineering infrastructure be accurately design to avoid these consequences. This paper evaluates the performances of four probability distributions; GEV, EV1, LP3 and P3 using the annual maxima precipitation series of 26 years for Warri Metropolis obtained from Nigerian Meteorological Agency (NiMet). The strength and weakness of the four probability distributions were examined with the goodness of fit (GOF) module of Easyfit software which implemented Kolmogorov - Smirnov (KS) and Anderson - Darling (AD) tests at 5% significance level. The Easyfit software fitted the precipitation series data to the four probability distributions and ranked the four probability distributions across the fifteen rainfall durations. Results show that for both KS and AD tests, GEV distribution was found to be best-fit distribution and it was applied to the development of IDF curves in Warri Metropolis, Nigeria. Furthermore, the IDF values obtained were applied in the development of three-parameter IDF models for return periods of 10 - , 15 -, 20 -, 25 - , 50 -, and 100-years. The mean absolute error, Nash – Sutcliffe Efficiency (NSE) and Root Mean Square Error (RMSE) indices computed for the IDF models increase with increasing return periods. The IDF curves and models depicted the general attributes of IDF curves and models. This study could be of significant academic value and improvement to professional practice in the design of storm water drainage systems. Therefore, the developed IDF curves and models are recommended to the Warri Urban Authority for inclusion in her stormwater handbooks and manuals.

Statistical analysis of precipitation time series in Dobrudja region

Temporal characteristics of precipitation evolution in Dobrudja, a region situated in the Southeastern part of Romania, are analyzed in this article, using a data base of ten monthly series, collected in the period January 1965-December 2005. This paper describes different methods to detect the break points existence in order to detect changes in evolution of the monthly precipitation series. The study indicates a constant trend of precipitation before 2000 and an increasing one after 2000, in concordance with the predictions for this region.

Projecting Hydrological Responses to Climate Change Using CMIP6 Climate Scenarios for the Upper Huai River Basin, China

The influence of climate change on the regional hydrological cycle has been an international scientific issue that has attracted more attention in recent decades due to its huge effects on drought and flood. It is essential to investigate the change of regional hydrological characteristics in the context of global warming for developing flood mitigation and water utilization strategies in the future. The purpose of this study is to carry out a comprehensive analysis of changes in future runoff and flood for the upper Huai River basin by combining future climate scenarios, hydrological model, and flood frequency analysis. The daily bias correction (DBC) statistical downscaling method is used to downscale the global climate model (GCM) outputs from the sixth phase of the Coupled Model Intercomparison Project (CMIP6) and to generate future daily temperature and precipitation series. The Xinanjiang (XAJ) hydrological model is driven to project changes in future seasonal runoff under SSP245 and SSP585 scenarios for two future periods: 2050s (2031–2060) and 2080s (2071–2100) based on model calibration and validation. Finally, the peaks over threshold (POT) method and generalized Pareto (GP) distribution are combined to evaluate the changes of flood frequency for the upper Huai River basin. The results show that 1) GCMs project that there has been an insignificant increasing trend in future precipitation series, while an obvious increasing trend is detected in future temperature series; 2) average monthly runoffs in low-flow season have seen decreasing trends under SSP245 and SSP585 scenarios during the 2050s, while there has been an obvious increasing trend of average monthly runoff in high-flow season during the 2080s; 3) there is a decreasing trend in design floods below the 50-year return period under two future scenarios during the 2050s, while there has been an significant increasing trend in design flood during the 2080s in most cases and the amplitude of increase becomes larger for a larger return period. The study suggests that future flood will probably occur more frequently and an urgent need to develop appropriate adaptation measures to increase social resilience to warming climate over the upper Huai River basin.

A new GA-IDW approach for interpolating the precipitation

This article presents a new algorithm for detecting the Inverse Distance Weighting Algorithm parameter (IDW) using an evolutionary technique. The algorithm was applied to interpolate 51 series of maximum annual precipitation series. Comparisons of its results with those of IDW and the optimized OIDW (a version of IDW optimized with PSO) are provided. The best performances are those of the actual approach.

Documentary-based climate reconstructions in the Czech Lands 1501–2020 CE and their European context

Annual and seasonal temperature, precipitation and drought index (SPI, SPEI, Z-index, PDSI) series covering the Czech Lands territory (now the Czech Republic) over 520 years (1501–2020 CE) reconstructed from documentary data combined with instrumental observations were analysed herein. The temperature series exhibits a statistically significant increasing trend, rising from ~1890 and particularly from the 1970s; 1991–2020 represents the warmest and driest 30-year period since 1501 CE. While the long-term precipitation total fluctuations (and derived SPI fluctuations) remain relatively stable with annual and decadal variabilities, past temperature increases are the key factor affecting recent increasing dryness in the SPEI, Z-index and PDSI series. The seasonal temperature series represent a broad European area, while the seasonal precipitation series show lower spatial correlations. A statistical attribution analysis conducted utilizing regression and wavelet techniques confirmed the influence of covariates related to volcanic activity (prompting temporary temperature decreases, especially during summer) and the North Atlantic Oscillation (influential in all seasons except summer) in the Czech climate reconstructions. Furthermore, components tied to multidecadal variabilities in the northern Atlantic and northern Pacific were identified in the temperature and precipitation series and in the drought indices, revealing notable shared oscillations, particularly at periods of approximately 70–100 years.

Study on Precipitation Evolution Characteristics in Tongzhou District of Beijing in Recent 65 Years

Based on the 65a (1956–2020) precipitation series data of 11 rainfall stations and 5 surrounding rainfall stations in Tongzhou District, Beijing, the evolution characteristics of precipitation in Tongzhou District on spatial, interannual and intra annual scales are comprehensively analyzed using cumulative anomaly method, 5a moving average method and spectral analysis method, and the future change trend is predicted using ARIMA model. The results show that: 1) the annual average precipitation in Tongzhou District is higher in the middle and northwest and lower in the southwest, and the precipitation between June to August, accounts for more than 70% of the annual precipitation; 2) In general, the precipitation shows a fluctuating downward trend at the rate of -2.42 mm a-1, in which the precipitation in summer decreases at the rate of -2.68 mm a-1, while the precipitation in spring and autumn increases at the rates of 0.35 mm a-1 and 0.26 mm a-1 respectively; 3) The abrupt change of precipitation occurred in 1959 and 2000, which were 990.2mm and 239.4mm respectively; 4) There are 3∼8a and 14∼16a oscillation periods on the inter annual scale of precipitation, the prediction results of ARIMA model show that the precipitation will increase about 40 mm in the next five years.

Spatiotemporal Trends and Detection of Changes in Hydrological and Climatic Variables of Modjo River Watershed, Ethiopia

Understanding hydro-climatic trends in space and time is crucial for water resource planning and management, agricultural productivity and climate change mitigation of a region. This study examined the spatiotemporal variations in precipitation, reference evapotranspiration (ETo) and streamflow in a tropical watershed located in the central highlands of Ethiopia. Temporal trend implications were analyzed using the Mann-Kendall test, and Theil-Sen approach, whereas the inverse distance weighted interpolation method was applied for spatial trend variability analysis. The result showed that a significant decreasing trends in streamflow for the major rainy (Kiremt: Jun - Sept) season and annual time scales. At the same time, the annual and monthly ETo followed significantly increasing trends, but there has been a trendless time series for most of the months and annual mean precipitation series for the period 1986 - 2015. The study indicated that the spatial variability of annual and seasonal precipitation series decreased from north to south and west to east, while this was increased for ETo both for annual and seasonal time series over the study watershed. The contribution of rainfall and mean temperature to streamflow decline was insignificant. It is pointed out that river flow regime is weakly affected by climate changes, hence human activities are stronger in explaining the river flow trends of the watershed. Therefore, urgent calls on the needs for reducing human-induced impacts, and implementing appropriate watershed management, conservation measures and an efficient use of water resources.