Search

A traversing system with a fast responding, photocell light sensor was used to measure mean irradiance in corn (Zea maysL.) canopies. Data were recorded by digital and analog recorders to analyze the sampling problems with a moving sensor. Sampling rates of once every 4 seconds for a row‐planted crop and once every 8 seconds for a uniformly planted crop were required to yield satisfactory mean irradiance values from a fast responding sensor. In addition, sensors with time constants at least as long as 8 seconds could be used to measure mean irradiance. Therefore, we recommend that the sampling rate for a traversing sensor in a maize leaf canopy be at least as frequent as the time constant of the sensor, but in any case the sampling rate does not need to be any more frequent than once every 4 seconds.

This study was undertaken to provide information about hydraulic conductivity, percent large pores, and bulk density, their interrelationships, and the magnitude of the sampling error components involved over a wide range of soil conditions. Data from approximately 10,000 individual core samples from about 900 sites in 7 states, furnished in a cooperative project with the Soil Conservation Service, United States Department of Agriculture, were analyzed. Each sample was classified according to management, genetic unit (great soil group, high family, low family, and series), texture, and horizon. Mean values for hydraulic conductivity, percent large pores and bulk density, when classified by textural groups, showed a consistent decrease in magnitude with increases in silt and clay content. Mean values for the three properties, when classified by great soil groups, showed consistent differences between groups. Correlations between hydraulic conductivity and percent large pores was positive and comparatively consistent; correlations between hydraulic conductivity and bulk density were negative and generally of a low absolute value. The indicated conclusion was that bulk density, in itself, is a poor indicator of soil permeability. Variance components for among cores within sites and horizons, and between sites, after removal of all other classifiable variation, indicated that the between site component of the sampling error was 2 to 3 times larger than the within site component. These components are used to estimate “distinguishable permeability classes” for various sampling schemes, and by assuming various cost ratios, optimum sampling rates are estimated.

The mean and variance of an estimate V(τ) of autocorrelation φ(τ) are calculated. Tho estimate is derived from a single sample, of finite length, of the time-series which is assumed to be stationary and ergodic. It is found that, for values of τ which make φ(τ) appreciable, tho variance depends markedly on the magnitude of the 4th moment of the first distribution of the time-series. On the other hand, for values of τ which make φ(τ) negligible, the variance in nearly all cases falls off linearly with τ, reaching zero when τ equals the sample length.A simple explanation is also given of the well-known fact that the Fourier transform of V(τ) cloos not yield a good estimate of the spectral density.

This paper applies a previously derived equation for sampling error in hybrid systems to the case of hybrid computer systems which use first-order analog hold de vices. Explicit expressions for sampling error are ob tained for a general hybrid computer program and for several classes of programs corresponding to particular cases. The equations apply for most types of hybrid com puter operation including the case for which the digital computer is used to integrate one or more of the prob lem equations.The error equations are used to examine the effect of various allocations of computing operations, both in general terms and through use of a particular example. The results, while providing no new insight into gross effects, give quantitative relations which sharpen intui tive guidelines for programming a hybrid computer.The equations place in evidence the following facts which are not apparent from a qualitative study: (1) The major components of sampling error depend on the first power of the sampling period, making the hybrid computer a first-order numerical method. (2) Digital execution time has an effect on error which is similar to that of the sampling period, but is weighted twice as heavily. (3) A numerical integration method in the digi tal computer produces negligible error in relation to other error sources if it is of order 2 or higher.