Maribeth Price

Venus: Crater distribution and plains resurfacing models

Detailed analysis of the distribution of craters on Venus using Mth nearest neighbor analysis, coupled with models based upon surface morphology constraints, indicates that the hypothesis of complete spatial randomness (CSR) cannot be rejected, but is not a unique model of the observed crater distribution. Based on morphologic mapping, the extensive volcanic plains can be divided into four units that have a spread in age of the order of 0.5T (the mean surface age of the planet). This four-unit plains model, along with its derivatives, produce test statistics that indicate such models also cannot be rejected. Further, the probability of obtaining a result at least as extreme as the observed test statistic given that the null hypothesis (model corresponds to Venus) is true is lowest for the CSR model. There is no particular reason to pick a CSR model (along with its implications for catastrophic resurfacing) as a constraint on the evolution of Venus, and there are geological reasons to choose the multiage models. We find that we cannot distinguish statistically among models that have two, three, or four distinct production ages within the plains. However, the hypothesis that the variation in crater density within all of the plains is due to a single random process can be rejected for two reasons. First, the binomial probability that such a process could exist within each of the plains units is ≤0.05 except the smallest and youngest unit, PL1. Second, using a chi-squared statistic to test the hypothesis that four plains units have the same age gives a p value of 10−4, indicating confident rejection of the hypothesis. Thus CSR cannot be used as a constraint on models of resurfacing or planetary evolution of Venus because of the non-uniqueness in matching such a model to the observed crater distribution and the strong indication of distinct ages within the plains with a significant spread in age. Geological and geophysical constraints provide our best clues for understanding Venus.

A method to obtain large quantities of reference data

Project managers often struggle with the need of sufficient reference data to train and test for reliable classifications and budget concerns that restrain the amount of justifiable field data collection. For a forest study, we supplemented our 207 ground‐measured field sites with 4000 additional photo‐interpreted reference sites. We first used aerial photography to identify the extent of homogenous regions around field‐data sites and then picked additional reference points within these areas. This approach is based on the notion that similar‐appearing areas close to a measured vegetation plot will contain approximately the same mix and density of species as the known site. This resulted in clusters of additional data points around actual field locations. We avoided overestimating the classification accuracy due to spatial autocorrelation by using an entire cluster of reference points exclusively as training or test data.

Impact of reference datasets and autocorrelation on classification accuracy

Reference data and accuracy assessments via error matrices build the foundation for measuring success of classifications. An error matrix is often based on the traditional holdout method that utilizes only one training/test dataset. If the training/test dataset does not fully represent the variability in a population, accuracy may be over – or under – estimated. Furthermore, reference data may be flawed by spatial errors or autocorrelation that may lead to overoptimistic results. For a forest study we first corrected spatially erroneous ground data and then used aerial photography to sample additional reference data around the field-sampled plots (Mannel et al. 2006). These reference data were used to classify forest cover and subsequently determine classification success. Cross-validation randomly separates datasets into several training/test sets and is well documented to perform a more precise accuracy measure than the traditional holdout method. However, random cross-validation of autocorrelated data may overestimate accuracy, which in our case was between 5% and 8% for a 90% confidence interval. In addition, we observed accuracies differing by up to 35% for different land cover classes depending on which training/test datasets were used. The observed discrepancies illustrate the need for paying attention to autocorrelation and utilizing more than one permanent training/test dataset, for example, through a k-fold holdout method.1 Now at: Cottey College, 6000 W. Austin, Nevada, MO 64772, USA.

Comparison of Combinations of Sighting Devices and Target Objects for Establishing Circular Plots in the Field

Many aspects of ecological research require measurement of characteristics within plots. Often, the time spent establishing plots is small relative to the time spent collecting and recording data. However, some studies require larger numbers of plots, where the time spent establishing the plot is consequential to the field effort. In open habitats, circular plots are easily established using a rope or tape. In tall or dense vegetation, however, considerable time can be spent ensuring that measures of plot radii are straight-line measurements. To rapidly establish fixed-radius plots in the field, common forest survey techniques can be used with a target object calibrated to the desired size of the plot. Although mentioned in past publications, the accuracy of establishing plots with these methods has not been evaluated. We tested the accuracy and precision in establishing fixed-radius plots using different sighting device/target object combinations. A laser rangefinder aimed at 10.2-cm PVC pipe was most accurate and precise, but expensive, and required careful handling. Wedge prisms used with a 10.2-cm PVC pipe or cylinder were accurate, precise, inexpensive, and easy to use.

Comparing Classification Results of Multi-Seasonal TM against AVIRIS Imagery – Seasonality more Important than Number of Bands

much smaller than for example a Landsat TM scene, which usually covers about 20,000 km. Vegetation studies can take advantage of the continuously available relectance; for example, NiemaNN et al. (2002) note that narrow spectral bands are necessary to detect some forest related parameters whose spectral range may be small. Hyperspectral data have been used to map land cover types such as woody vegetation (Wylie et al. 2000, UstiN &Xiao 2001), leafy spurge (Williams&HUNt Jr. 2002), shrub recovery after ire (riaNo et al. 2002), vegetation in semi arid ecosystems (asNer& HeidebrecHt 2002, okiN et al. 2001) or lake water quality (HoogeNboom et al. 1998, tHiemaNN & kaUfmaNN 2002). The nearly continuous spectrum also has its costs, in an economical, computational, and spatial sense. A typical AVIRIS scene holds

C-Lock-A Method to Maximize Carbon Sequestration Value to Agro-forestry Producers and Purchasers

Publisher Summary This chapter provides a detailed description of C-Lock, a patent-pending Web-based carbon sequestration accounting and marketing tool. C-Lock, developed by Dr. Zimmerman and his colleagues at the South Dakota School of Mines and Technology, aggregates carbon emission reduction offsets from individual land parcels and prepares certified units for sale in the marketplace. The C-Lock process allows agricultural producers to quantify the impact of specific land-use management practices for specific agricultural land parcels on the sequestration of carbon in soil and vegetation. It also aggregates carbon emission reduction offsets for individual land parcels into units that can be efficiently marketed. This comprehensive tool has been designed to serve as an interface to link agricultural producers, carbon sequestration science and policy, and those who wish to purchase carbon emission reduction offsets. The C-Lock process incorporates three levels of data validation. Level I validation compares producer input data with lookup data for regional land use. Level II verification consists of a random audit to compare satellite data for a land parcel with reported data. Level III verification consists of submitting all data to a third party to operate the model and confirm the results. The C-Lock system must account for the variability in carbon accounting that results from a wide range of sources.