Land cover dynamics monitoring with Landsat data in Kunming,China: a cost-effective sampling and modelling scheme using Google Earth imagery and random forests

Changes in forest composition impact ecological services, and are considered important factors driving global climate change. A hybrid sampling method along with a modelling approach to map current and past land cover in Kunming, China is reported. MODIS land cover (2001–2011) data-sets were used to detect pixels with no apparent change. Around 3000 ‘no change points’ were systematically selected and sampled using Google Earth’s high-resolution imagery. Thirty-five per cent of these points were verified and used for training and validation. We used Random forests to classify multi-temporal Landsat imagery. Results show that forest cover has had a net decrease of 14385?ha (1.3% of forest area), which was primary converted to shrublands (11%), urban and barren land (2.7%) and agriculture (2.5%). Our validation indicates an overall accuracy (Kappa) of 82%. Our methodology can be used to consistently map the dynamics of land cover change in similar areas with minimum costs.     

Ground-based mid infra-red observations of nearby starburst and AGN galaxies

We present preliminary results of recent spatially resolved mid infrared (Mid-IR) spectroscopy with the MICHELLE instrument on UKIRT of NGC 7469, observed as part of a larger programme to study a sample of starburst and AGN in the Mid-IR. This revised version was published online in August 2006 with corrections to the Cover Date.     

Remobilization of authigenic uranium in marine sediments by bioturbation

Uranium behaves as a nearly conservative element in oxygenated seawater, but it is precipitated under chemically reducing conditions that occur in sediments underlying low-oxygen bottom water or in sediments receiving high fluxes of particulate organic carbon. Sites characterized by a range of bottom-water oxygen (BWO) and organic carbon flux (OCF) were studied to better understand the conditions that determine formation and preservation of authigenic U in marine sediments. Our study areas are located in the mid latitudes of the northeast Pacific and the northwest Atlantic Oceans, and all sites receive moderate (0.5 g/cm² kyr) to high (2.8 g/cm² kyr) OCF to the sediments. BWO concentrations vary substantially among the sites, ranging from <3 to ∼270 μM. A mass balance approach was used to evaluate authigenic U remobilization at each site. Within each region studied, the supply of particulate nonlithogenic U associated with sinking particles was evaluated by means of sediment traps. The diffusive flux of U into sediments was calculated from pore-water U concentration profiles. These combined sources were compared with the burial rate of authigenic U to assess the efficiency of its preservation. A large fraction (one-third to two-thirds) of the authigenic U precipitated in these sediments via diffusion supply is later regenerated, even under very low BWO concentrations (∼15 μM). Bioturbating organisms periodically mix authigenic U-containing sediment upward toward the sediment-water interface, where more oxidizing conditions lead to the remobilization of authigenic U and its loss to bottom waters.     

Decoupled elastic prestack depth migration

This paper presents a new decoupled form of the formula for common-shot or common-receiver amplitude-preserving elastic prestack depth migration (PreSDM), which can be used for estimating angle-dependent elastic reflection coefficients in laterally inhomogeneous anisotropic media. The multi-shot or multi-receiver extension of this formula is suitable for automated prestack amplitude-versus-angle (AVA) elastic inversion of ocean-bottom cable (OBC), walkaway VSP (WVSP) or standard towed-cable data at any subsurface location. The essence of the theory is a systematic application of the stationary-phase principle and high-frequency approximations to the basic elastic Green's theorem. This leads to nonheuristic explicit wave mode decoupling and scalarization of vector PreSDM. Used in combination, ray-trace and finite-difference (FD) eikonal solvers create a useful tool to calculate accurate Green's function travel time and amplitude maps. Examples of synthetic OBC data and applications to field WVSP data show that the new imaging technique can produce a clear multi-mode elastic image.     

Mineralogy and petrology of Al-rich objects and amoeboid olivine aggregates in the CH carbonaceous chondrite North West Africa 739

The aluminum-rich (>10 wt% Al₂O₃) objects in the CH carbonaceous chondrite North West Africa (NWA) 739 include Ca,Al-rich inclusions (CAIs), Al-rich chondrules, and isolated mineral grains (spinel, plagioclase, glass). Based on the major mineralogy, 54 refractory inclusions found in about 1 cm² polished section of NWA 739 can be divided into hibonite-rich (16%), grossite-rich (26%), melilite-rich (28%), spinel-pyroxene-rich (16%) CAIs, and amoeboid olivine aggregates, (AOA's, 17%). Most CAIs are rounded, 25–185 μm (average=70 μm) in apparent diameter, contain abundant, tiny perovskite grains, and typically surrounded by a single- or double-layered rim composed of melilite and/or Al-diopside; occasionally, layers of spinel+hibonite and forsterite are observed. The AOAs are irregularly shaped, 100–250 μm (average=175 μm) in size, and consist of forsterite, Fe,Ni-metal, and CAIs composed of Al-diopside, anorthite, and minor spinel. One AOA contains compact, rounded melilite-spinel-perovskite CAIs and low-Ca pyroxene replacing forsterite. The Al-rich (>10 wt% bulk Al₂O₃) chondrules are divided into Al-diopside-rich and plagioclase-rich. The Al-diopside-rich chondrules, 50–310 μm (average=165 μm) in apparent diameter, consist of Al-diopside, skeletal forsterite, spinel, ±Al-rich low-Ca pyroxene, and ±mesostasis. The plagioclase-rich chondrules, 120–455 μm (average=285 μm) in apparent diameter, are composed of low-Ca and high-Ca pyroxenes, forsterite, anorthitic plagioclase, Fe,Ni-metal nodules, and mesostasis. The isolated spinel occurs as coarse, 50–125 μm in size, subhedral grains, which are probably the fragments of Al-diopside chondrules. The isolated plagioclase grains are too coarse (60–120 μm) to have been produced by disintegration of chondrules or CAIs; they range in composition from nearly pure anorthite to nearly pure albite; their origin is unclear. The Al-rich objects show no evidence for Fe-alkali metasomatic or aqueous alteration; the only exception is an Al-rich chondrule fragment with anorthite replaced by nepheline. They are texturally and mineralogically similar to those in other CH chondrites studied (Acfer 182, ALH85085, PAT91467, NWA 770), but are distinct from the Al-rich objects in other chondrite groups (CM, CO, CR, CV). The CH CAIs are dominated by very refractory minerals, such as hibonite, grossite, perovskite and gehlenitic melilite, and appear to have experienced very low degrees of high-temperature alteration reactions. These include replacement of grossite by melilite, of melilite by anorthite, diopside, and spinel, and of forsterite by low-Ca pyroxene. Only a few CAIs show evidence for melting and multilayered Wark-Lovering rims. These observations may suggest that CH CAIs experienced rather simple formation history and escaped extensive recycling. In order to preserve the high-temperature mineral assemblages, they must have been efficiently isolated from the hot nebular region, like some chondrules and the zoned Fe,Ni-metal grains in CH chondrites.     

An easy approach to assess the susceptibility of a landslide by utilizing simple raster overlay methods: A case study on Huangtupo landslide (P.R. China)

Semi qualitative index based methods using rankings and ratings are commonly used in susceptibility estimations over a wide area. However, generalized ranking and ratings are not applicable for one single landslide. This paper gives an easy and transferable approach to a susceptibility assessment of Huangtupo landslide (P.R. China), using raster addition without taking account for ranking and ratings. Slope, aspect, curvature, location and drainage buffer distance raster data sets have been obtained out of open source digital elevation models using ESRI’s ArcGIS. These conditioning factor raster data sets have been translated into raster data sets including simple yes or no criteria, referring to triggering or not. Subsequently they have been added by raster math to acquire a simple raster overlay map. After that this map is compared to initial displacement measurements, obtained by using a ground based synthetic aperture radar device. Acquired data is recalculated to a raster data set using the same spatial extent, to provide the possibility of comparison of the two raster data sets. The results reveal, that 76.35% of all measured movements occur in areas where raster cells include three or more conditioning factors, indicating that easy raster math operations can lead to satisfying results in local scale.     

Office dispersal in metropolitan areas II: Case study results and conclusions

A case study of office dispersal options in Sydney, Australia is presented within a Planning Balance Sheet Framework. Two dispersal options, one involving a limited number of large subcentres, the other a large number of smaller subcentres are examined. It is shown that both options are likely to result in considerable work-journey time savings for employees in relocated offices (particularly clerical and female workers) and considerable improvements in job access for many suburban residents. Overall, however, the large subcentre option appears to offer greater net benefits to both office employees and relocated organizations. Yet both options are likely to generate high external costs in terms of suburban road congestion, increased energy consumption, financial losses to public transport operators, and inequities of access to captive public transport users. These problems could be avoided if the government were to embark on a programme of upgrading public transport services and restricting car usage to the point where the majority of suburban commuters used public transport. This would involve considerable public expenditure, which is unlikely to be forthcoming in the present economic climate. But without such a programme, a policy of office dispersal would create more problems than it would solve.     

Longwave Measurements for the Coast of British Columbia and Improvements to the Tsunami Warning Capability

A few years ago the Canadian Hydrographic Service initiated a major upgrade toall tide gauges and tsunami stations on the coast of British Columbia (B.C.). Thisprogram was undertaken to address shortcomings of the earlier digital systems andwas driven by concerns about emergency response continuity in the year 2000. By1999, thirteen tide gauge stations had been installed and were operational. Three ofthese stations (Tofino, Winter Harbour, and Langara) were selected for use as tsunamiwarning stations. Several years of continuous, high quality data have now been collectedat these stations and used for analysis of long waves in the tsunami frequency band.Careful examination of these data revealed two weak tsunamis recorded by severalB.C. stations: a distant tsunami of June 23, 2001 generated by the Peru Earthquake(M_w = 8.4), and a local tsunami of October 12, 2001 induced by the Queen Charlotte Earthquake (M_w = 6.3$). Spectral characteristics of these two tsunamis are compared with the spectral characteristics of long waves generated by a strong storm (October, 2000) and of ordinary background oscillations. The topographic admittance functions (frequency responses) constructed for all stations showed that most of them (in particular, Winter Harbour, Tofino, Bamfield, Port Hardy, and Victoria) have strong resonance at periods from 2.5 to 20 min, indicating that these locations are vulnerable to relatively high-frequency tsunamis. The Winter Harbour station also has two strong resonant peaks with periods of 30 and 47 min and with amplification factors of about 7. The estimated source functions show very clear differences between long waves associated with the seismic source (typical periods 10–30 min) and those generated by a storm, which typically have shorter periods and strong energy pumping from high-frequencies due to non-linear interaction of wind waves.