Mohsen Nasseri

Forecasting monthly urban water demand using Extended Kalman Filter and Genetic Programming

Research highlights? We implemented GP for developing suitable functional forms for water demand forecasting. ? These formula have been evaluated by mathematical sensitivity analysis and the best one has been chosen. ? Then, EKF as a nonlinear data assimilator has been used for increasing accuracy of the best model result. In this paper, a hybrid model which combines Extended Kalman Filter (EKF) and Genetic Programming (GP) for forecasting of water demand in Tehran is developed. The initial goal of the current work is forecasting monthly water demand using GP for achieving an explicit optimum formula. In the proposed model, the EKF is applied to infer latent variables in order to make a forecasting based on GP results of water demand. The available dataset includes monthly water consumption of Tehran, the capital of Iran, from 1992 to 2002. Five best formulas based on GP results on this dataset are presented. In these models, the first five to three lags of observed water demand are used as probable and independent inputs. For each model, sensitivity of the results for each input is measured mathematically. A model with the most compatibility of the computed versus the observed water demand is used for filtering based on EKF method. Results of GP and hybrid models of EKFGP demonstrate the visible effect of observation precision on water demand prediction. These results can help decision makers of water resources to reduce their risks of online water demand forecasting and optimal operation of urban water systems.

The use of a genetic algorithm-based search strategy in geostatistics: application to a set of anisotropic piezometric head data

In any geostatistical study, an important consideration is the choice of an appropriate, repeatable, and objective search strategy that controls the nearby samples to be included in the location-specific estimation procedure. Almost all geostatistical software available in the market puts the onus on the user to supply search strategy parameters in a heuristic manner. These parameters are solely controlled by geographical coordinates that are defined for the entire area under study, and the user has no guidance as to how to choose these parameters. The main thesis of the current study is that the selection of search strategy parameters has to be driven by data-both the spatial coordinates and the sample values-and cannot be chosen beforehand. For this purpose, a genetic-algorithm-based ordinary kriging with moving neighborhood technique is proposed. The search capability of a genetic algorithm is exploited to search the feature space for appropriate, either local or global, search strategy parameters. Radius of circle/sphere and/or radii of standard or rotated ellipse/ellipsoid are considered as the decision variables to be optimized by GA. The superiority of GA-based ordinary kriging is demonstrated through application to the Wolfcamp Aquifer piezometric head data. Assessment of numerical results showed that definition of search strategy parameters based on both geographical coordinates and sample values improves cross-validation statistics when compared with that based on geographical coordinates alone. In the case of a variable search neighborhood for each estimation point, optimization of local search strategy parameters for an elliptical support domain-the orientation of which is dictated by anisotropic axes-via GA was able to capture the dynamics of piezometric head in west Texas/New Mexico in an efficient way.

COMPARISON BETWEEN ACTIVE LEARNING METHOD AND SUPPORT VECTOR MACHINE FOR RUNOFF MODELING

Comparison Between Active Learning Method and Support Vector Machine for Runoff Modeling In this study Active Learning Method (ALM) as a novel fuzzy modeling approach is compared with optimized Support Vector Machine (SVM) using simple Genetic Algorithm (GA), as a well known datadriven model for long term simulation of daily streamflow in Karoon River. The daily discharge data from 1991 to 1996 and from 1996 to 1999 were utilized for training and testing of the models, respectively. Values of the Nash-Sutcliffe, Bias, R2, MPAE and PTVE of ALM model with 16 fuzzy rules were 0.81, 5.5 m3 s-1, 0.81, 12.9%, and 1.9%, respectively. Following the same order of parameters, these criteria for optimized SVM model were 0.8, -10.7 m3 s-1, 0.81, 7.3%, and -3.6%, respectively. The results show appropriate and acceptable simulation by ALM and optimized SVM. Optimized SVM is a well-known method for runoff simulation and its capabilities have been demonstrated. Therefore, the similarity between ALM and optimized SVM results imply the ability of ALM for runoff modeling. In addition, ALM training is easier and more straightforward than the training of many other data driven models such as optimized SVM and it is able to identify and rank the effective input variables for the runoff modeling. According to the results of ALM simulation and its abilities and properties, it has merit to be introduced as a new modeling method for the runoff modeling. Porovnanie Metódy Aktívneho Učenia S Metódou Vektormi Podporených Strojov Pri Modelovaní Odtoku Cieľom štúdie bolo porovnať možnosti dlhodobej simulácie denných prietokov v rieke Karoon pomocou novovyvinutej fuzzy metódy aktívneho učenia (Active Learning Method - ALM) a známej metódy vektormi podporených strojov (Support Vector Machine - SVM), optimalizovanej genetickým algoritmom (GA). Na tréning a testovanie modelov boli použité časové rady denných prietokov za obdobie rokov 1991 až 1996 a 1996 až 1999. Hodnoty parametrov Nash-Sutcliffe, Bias, R2, MPAE a PTVE pre model ALM boli 0,81; 5,5 m3 s-1; 0,81; 12,9% a 1,9%. Parametre v tom istom poradí pre model SVM boli 0,8 -10,7 m3 s-1, 0,81; 7,3%; a -3,6%. Z výsledkov simulácií vyplýva, že aplikáciou metód ALM a SVM možno získať porovnateľné a akceptovateľné výsledky. Podobnosť výsledkov medzi ALM a SVM implikuje vhodnosť novovyvinutej metódy ALM pre simuláciu odtoku. Tréning ALM je ľahší a jednoduchší ako je tréning ďalších dátami riadených modelov podobného typu. Navyše algoritmus ALM je schopný identifikovať a zoradiť efektívne vstupné premenné pre modelovanie odtoku. Na základe dosiahnutých výsledkov možno metódu ALM zaradiť medzi nové, alternatívne metódy modelovania odtoku.

Backcasting long-term climate data: evaluation of hypothesis

Most often than not, incomplete datasets or short-term recorded data in vast regions impedes reliable climate and water studies. Various methods, such as simple correlation with stations having long-term time series, are practiced to infill or extend the period of observation at stations with missing or short-term data. In the current paper and for the first time, the hypothesis on the feasibility of extending the downscaling concept to backcast local observation records using large-scale atmospheric predictors is examined. Backcasting is coined here to contrast forecasting/projection; the former is implied to reconstruct in the past, while the latter represents projection in the future. To assess our hypotheses, daily and monthly statistical downscaling models were employed to reconstruct past precipitation data and lengthen the data period. Urmia and Tabriz synoptic stations, located in northwestern Iran, constituted two case study stations. SDSM and data-mining downscaling model (DMDM) daily as well as the group method of data handling (GMDH) and model tree (Mp5) monthly downscaling models were trained with National Center for Environmental Prediction (NCEP) data. After training, reconstructed precipitation data of the past was validated against observed data. Then, the data was fully extended to the 1948 to 2009 period corresponding to available NCEP data period. The results showed that DMDM performed superior in generation of monthly average precipitation compared with the SDSM, Mp5, and GMDH models, although none of the models could preserve the monthly variance. This overall confirms practical value of the proposed approach in extension of the past historic data, particularly for long-term climatological and water budget studies.

Spatial Scale Resolution of Prognostic Hydrological Models: Simulation Performance and Application in Climate Change Impact Assessment

In this paper, long-term hydrological response of a watershed to climate change was investigated taking into account the spatial scale effect on the performance of hydrological models. A water balance model was used in which variations of soil moisture, snow budget, deep infiltration and interactions with groundwater resources were modeled. Four various combinations of sub-catchment delineation, altitudinal discretization and division into square-shaped grids were tested for semi-distributed water balance modeling of a basin located in southwest of Iran, namely Roodzard Basin, with arid and semiarid climate based on Köppen-Geiger climate classification. The results showed improvement in the model performances when spatial variations of the meteorological data and topographic characteristics of the basin were incorporated in the modeling process. The effects of spatial scale resolution dependency were evaluated in projecting streamflow for various climate change scenarios. The results showed that finer resolution of grid cells in the semi-distributed model does not necessarily result in more accurate estimation of monthly streamflows and altitudinal discretization provides almost same accuracy as the results of grid-based models. Moreover, probability distribution of projections obtained from water balance models for A2 and B2 of Special Report on Emissions Scenarios (SRES) scenarios presented less coefficient of variation and skewness compared with historical observations.