Dam failure peak outflow prediction through GEP-SVM meta models and uncertainty analysis

Accurate prediction of a breached dam's peak outflow is a significant factor for flood risk analysis. In this study, the capability of Support Vector Machine and Kernel Extreme Learning Machine as kernel-based approaches and Gene Expression Programming method was assessed in breached dam peak outflow prediction. Two types of modeling were considered. First, only dam reservoir height and volume at the failure time were used as the input combinations (state 1). Then, soil characteristics were added to input combinations to investigate particularly the impact of soil characteristics (state 2). Results showed that the use of only soil characteristics did not lead to a desired accuracy; however, adding soil characteristics to input combinations (state 2) improved the models' accuracy up to 40%. The outcome of the applied models was also compared with existing empirical equations and it was found the applied models yielded better results. Sensitivity analysis results showed that dam height had the most important role in the peak outflow prediction, while the strength parameters did not have significant impacts. Furthermore, for assessing the best-applied model dependability, uncertainty analysis was used and the results indicated that the SVM model had an allowable degree of uncertainty in peak outflow modelling.

Performance of two small experimental bioretention cells during the first year of operation

<p>Due to increasing urbanization, bioretention cells are becoming an increasingly popular solution for stormwater management. The data on long term performance bioretention are still sparse. The aim of this study was to set-up two experimental bioretention cells designed for long-term monitoring and to evaluate the rainfall-runoff characteristics, assess the development of properties of the biofilter, and dynamics of plant growth during the first growing season.</p><p>Two identical experimental bioretention cells were established. The first collects water from the roof and the second is supplied from the tank for simulating artificial rainfall. The 30 cm thick biofilter soil mixture is composed of 50% sand, 30% compost, and 20% topsoil. Bioretention cells are isolated from the surrounding soil by a waterproof membrane. Both bioretention cells are instrumented by an identical system of sensors. Four time-domain reflectometry probes monitor soil water contents 20 cm below the surface. Five tensiometers record the water potential in a biofilter. The amount of a discharge from each bioretention cell is determined by a tipping bucket flowmeter. A ponding depth is recorded by an ultrasonic sensor.</p><p>Rainfall-runoff episodes were evaluated for the period from 18.6. 2018 to the 22.11.2018. 17 episodes were evaluated for bioretention cell with the inflow of stormwater from the roof. Six ponding experiments were done in the bioretention cell with an artificial supply. Rainfall depth, maximal rainfall intensity, episode duration, runoff coefficient, and maximal peak outflow rate from both bioretention cells were determined for each episode. The effective saturated hydraulic conductivity was determined using Darcy’s law under the assumption of one-dimensional, vertical flow. The estimation method was verified by simulating two-dimensional variably saturated flow using HYDRUS-2D. Outflow water quality was measured from one bioretention cell during ponding experiments.</p><p>The runoff coefficient for the entire period of the growing season was 0.72, while the peak outflow reduction for individual rainfall events ranged between 75% to 95% for the bioretention cell connected to the roof. The runoff coefficient determined from artificial ponding events was 0.86 for the event started in the partially saturated biofilter, while it was nearly 1.0 for all subsequent artificial ponding events. The peak flow reduction ranged from 19% to 30%. The saturated hydraulic conductivity of biofilter with a natural rainfall supply ranged between 1.6·10<sup>-6</sup> to 8.6∙10<sup>-6</sup> m·s<sup>-1</sup>, which is significantly less than hydraulic conductivity 1.3∙10<sup>-4</sup> m·s<sup>-1</sup> measured in the laboratory on packed samples. Perennials Aster, Hemerocallis and Molinia have shown good growth and adaptation to conditions in bioretention cells. In the case of the current experiment, the gravel mulch layer has proven to be an effective barrier to reducing evaporation. The values of total suspended solids and turbidity were highly correlated and generally high, especially at the beginning of outflow in artificial ponding experiments. The value of electrical conductivity reached up to 2200 µS·cm<sup>-1</sup>, this may be due to the higher compost content in the soil. The monitoring of bioretention cells continues in order to record long term changes in the performance of the bioretention cells.</p>

Hubble Space Telescope Observations of [O III] Emission in Nearby QSO2s: Physical Properties of the Ionised Outflows

We use Hubble Space Telescope (HST)/ Space Telescope Imaging Spectrograph (STIS) long-slit G430M and G750M spectra to analyse the extended [O III] λ5007 emission in a sample of twelve nearby (z >0.12) luminous (Lbol > 1.6 × 1045 erg s−1) QSO2s. The purpose of the study is to determine the properties of the mass outflows of ionised gas and their role in AGN feedback. We measure fluxes and velocities as functions of radial distances. Using Cloudy models and ionising luminosities derived from [O III] λ5007, we are able to estimate the densities for the emission-line gas. From these results, we derive masses of [O III]-emitting gas, mass outflow rates, kinetic energies, kinetic luminosities, momenta and momentum flow rates as a function of radial distance for each of the targets. For the sample, masses are several times 103M⊙ − 107M⊙ and peak outflow rates are 9.3 × 10−3M⊙ yr−1 to 10.3 M⊙ yr−1. The peak kinetic luminosities are 3.4 × 10−8 to 4.9 × 10−4 of the bolometric luminosity, which does not approach the 5.0 × 10−3 - 5.0 × 10−2 range required by some models for efficient feedback. For Mrk 34, which has the largest kinetic luminosity of our sample, in order to produce efficient feedback there would have to be 10 times more [O III]-emitting gas than we detected at its position of maximum kinetic luminosity. Three targets show extended [O III] emission, but compact outflow regions. This may be due to different mass profiles or different evolutionary histories.