The Initial Abstraction ratio (Ia/S, or λ) in the Curve Number (CN) method was assumed in its original development to have a value of 0.20. Using event rainfall-runoff data from several hundred plots this assumption is investigated, and λ values determined by two different methods. Results indicate a λ value of about 0.05 gives a better fit to the data and would be more appropriate for use in runoff calculations. The effects of this change are shown in terms of calculated runoff depth and hydrograph peaks, CN definition, and in soil moisture accounting. The effect of using λ=0.05 in place of the customary 0.20 is felt mainly in calculations that involve either lower rainfall depths or lower CNs.

We propose a reversible reaction mechanism with a single stationary state in which certain concentrations assume either high or low values dependent on the concentration of a catalyst. The properties of this mechanism are those of a McCulloch-Pitts neuron. We suggest a mechanism of interneuronal connections in which the stationary state of a chemical neuron is determined by the state of other neurons in a homogeneous chemical system and is thus a "hardware" chemical implementation of neural networks. Specific connections are determined for the construction of logic gates: AND, NOR, etc. Neural networks may be constructed in which the flow oftime is continuous and computations are achieved by the attainment of a stationary state of the entire chemical reaction system, or in which the flow of time is discretized by an oscillatory reaction. In another article, we will give a chemical implementation of finite state machines and stack memories, with which in principle the construction of a universal Turing machine is possible.Computations may be supported by many different systems (1, 2), including physical systems like the digital computer, Fredkin logic gates (3), billiard-ball collisions (4), enzymes operating on a polymer chain (1, 5), and more abstract systems like cellular automata (6-8), partial differential equations that simulate cellular automata (9), generalized shifts (4), and neural networks (10-13). Some of these systems can be computationally universal and thus are formally equivalent with a universal Turing machine (10,14). We may inquire about whether computationally universal devices may be constructed solely from chemical reaction mechanisms in a homogeneous medium. All living entities process information to varying degrees, and this can occur only by chemical means. It is for this reason alone that the subject is ofinterest. In this article, we discuss the construction of chemical networks where coupled reaction mechanisms implement "programmed" computations as the concentrations evolve in time. It has already been noted that bistable chemical systems are in many ways analogous to a flip-flop circuit, by coupling bistable reactions it is possible to build universal automata (15,16), and that various chemical mechanisms share a formal relationship with electronic devices (17, 18). We address the construction of computational devices from the viewpoint of neural networks. We propose a chemical reaction network, which is a "hardware" implementation of a neural network, and hence the network can in principle be as powerful as a universal Turing machine (10).Neural networks are a versatile basis for computation (19). Any finite state machine, and hence the finite state part of a universal Turing machine, can be simulated by a neural network (10,20). Neural networks also form the basis of many collective computational systems such as feedforward networks or Hopfield's network (11-13). A chemical neural network may serve as the "hardware" for any of the approaches to computation. We present hardw...

No abstract

No abstract

No abstract

Iwara A.I. (2014): Evaluation of the variability in runoff and sediment loss in successional fallow vegetation of Southern Nigeria. Soil & Water Res., 9: 77-82. The effects of three different ages of natural fallow vegetation on runoff and sediment loss were investigated in a part of the rainforest zone of Nigeria. Measurements of runoff amount and sediment loss were made for the months of March to November in 2012 rainy season using runoff plots of 40 m 2. The average runoff amount for the 5-year-old, 3-year-old, and farmland plots were 0.47, 0.26, and 0.41 mm respectively. The average sediment loss on the 5-year-old, 3-year-old, and farmland plots were 209.24, 50.54, and 124.68 kg/ha, respectively. The lowest losses for both runoff and sediment were recorded on the 3-year-old plot, while the 5-year-old plot experienced the highest losses. The variations in runoff and sediment loss among the treatments were significant at P < 0.001. The results evidently showed that rainfall was principally responsible for the erosional losses on all the fallow treatments, and that ground cover (density of herbs) and girth helped to reduce sediment loss on the 3-year-old and farmland surfaces, respectively. The high amount of erosional losses experienced on the 5-year-old fallow than on the 3-year-old fallow and farmland plots imply that fallow that is not adequately protected by ground cover experiences accelerated soil erosion. The continuous loss in topsoil rich in plant nutrients may prolong the optimal capacity of the soil to regain its loss nutrient for subsequent food crop cultivation.

Experiments on pattern recognition are performed with a network of eight open, bistable, mass-coupled chemical reactors. A programming rule is used to determine the network connectivity in order to store sets of stationary patterns of reactors with low or high concentrations. Experiments show that these stored patterns can be recalled from similar initial patterns. To our knowledge, this is the first chemical implementation of a type of neural network computing device. The experiments on this small network agree with simulations and support the predictions of the performance of large networks.In a series of articles'-6 we have described the theoretical implementation of logic functions and both sequential and parallel computations by macroscopic chemical kinetics. In two of these articles1.2 we described a spatially distributed chemical network capable of recalling stored patterns from related inputs. The network is based on ideas taken from neural network t h e~r y .~-I~ In this letter, we report on experiments with a pattem recognition device of this type. The computation process is carried out by a chemical reaction with multiple steady states in a network of continuous flow stirred tank reactors (CSTRs). We have studied a network of eight CSTRs coupled by mass transport, in this case reciprocal mass pumping. Each reactor can exist in one of two stable steady states, and the set of states of the reactors constitutes a pattem; hence each reactor is a pixel in all of the patterns. With a programming rule, arbitrary sets of patterns can be stored in the network. Our chemical network uses bistable elements whereas typical neural networks use monostable McCullouch-Pitts type neurons." There are many physical, chemical, and biological processes which exhibit bistability. One of the best studied bistable chemical reactions is the iodate-arsenous acid reaction @ Abstract published in Advance ACS Abstracts, June 1, 1995.0022-3654/95/2099-10063$09.00/0 run in a CSTR.'2,13 When iodate is in stoichiometric excess, the overall reaction is 210,-+ SH,AsO, + 2H' -I, + SH,AsO, + H,O (1) At low flow rates, only a high iodine state exists, and vice versa. At intermediate flow rates, autocatalysis in iodide results in bistable states of either low or high iodine concentration. The state of each reactor is determined visually by adding starch to the reactor inflows, the high iodine state being blue and the low iodine state colorless. In our experiments each reactor is fed with an identical flow of reagents, at a flow rate corresponding to a point within the bistability limits (see Figure 1).To network the reactors, the CSTRs are connected by mass transport. Mass transport is implemented by reciprocal mass pumping using two 16-channel peristaltic pumps. Both pumps operate at identical flow rates, thereby ensuring that there is no net mass transfer between reactors. It can be shown that systems of this type possess only steady-state attractors.I6 The network is programmed by setting the pumping strengths between every pair of CSTRs ...

With methods developed in a prior article on the chemical kinetic implementation of a McCulloch-Pitts neuron, connections among neurons, logic gates, and a clocking mechanism, we construct examples of clocked rmite-state machines. These machines include a binary decoder, a binary adder, and a stack memory. An example of the operation of the binary adder is given, and the chemical concentrations corresponding to the state of each chemical neuron are followed in time. Using these methods, we can, in principle, construct a universal Turing machine, and these chemical networks inherit the halting problem.In a prior article (1) and rapid equilibration occurs only during the short time interval when the concentration of E is large. In Fig. 1 we show schematically the time variation of the concentrations of Ai and A, as determined by the concentration of Ci. Ai is the state of neuron i at a given time, say t = 0 for the interval 0 to 1 in Fig. 1 and determines