Baojuan Zheng

A support vector machine to identify irrigated crop types using time-series Landsat NDVI data

Abstract Site-specific information of crop types is required for many agro-environmental assessments. The study investigated the potential of support vector machines (SVMs) in discriminating various crop types in a complex cropping system in the Phoenix Active Management Area. We applied SVMs to Landsat time-series Normalized Difference Vegetation Index (NDVI) data using training datasets selected by two different approaches: stratified random approach and intelligent selection approach using local knowledge. The SVM models effectively classified nine major crop types with overall accuracies of >86% for both training datasets. Our results showed that the intelligent selection approach was able to reduce the training set size and achieved higher overall classification accuracy than the stratified random approach. The intelligent selection approach is particularly useful when the availability of reference data is limited and unbalanced among different classes. The study demonstrated the potential of utilizing multi-temporal Landsat imagery to systematically monitor crop types and cropping patterns over time in arid and semi-arid regions.

Measuring the spatial arrangement of urban vegetation and its impacts on seasonal surface temperatures

Urban forestry is an important component of the urban ecosystem that can effectively ameliorate temperatures by providing shade and through evapotranspiration. While it is well known that vegetation abundance is negatively correlated to land surface temperature, the impacts of the spatial arrangement (e.g. clustered or dispersed) of vegetation cover on the urban thermal environment requires further investigation. In this study, we coupled remote sensing techniques with spatial statistics to quantify the configuration of vegetation cover and its variable influences on seasonal surface temperatures in central Phoenix. The objectives of this study are to: (1) determine spatial arrangement of green vegetation cover using continuous spatial autocorrelation indices combined with high-resolution remotely-sensed data; (2) examine the role of grass and trees, especially their spatial patterns on seasonal and diurnal land surface temperatures by controlling the effects of vegetation abundance; (3) investigate the sensitivity of the vegetation–temperature relationship at varying geographical scales. The spatial pattern of urban vegetation was measured using a local spatial autocorrelation index—the local Moran’s Iv . Results show that clustered or less fragmented patterns of green vegetation lower surface temperature more effectively than dispersed patterns. The relationships between the local Moran’s Iv and surface temperature are evidenced to be strongest during summer daytime and lowest during winter nighttime. Results of multiple regression analyses demonstrate significant impacts of spatial arrangement of vegetation on seasonal surface temperatures. Our analyses of vegetation spatial patterns at varying geographical scales suggest that an area extent of ˜200 m is optimal for examining the vegetation–temperature relationship. We provide a methodological framework to quantify the spatial pattern of urban features and to examine their impacts on the biophysical characteristics of the urban environment. The insights gained from our study results have significant implications for sustainable urban development and resource management.

Does the spatial arrangement of urban landscape matter? Examples of urban warming and cooling in Phoenix and Las Vegas

Abstract This study examines the impact of spatial landscape configuration (e.g., clustered, dispersed) on land‐surface temperatures (LST) over Phoenix, Arizona, and Las Vegas, Nevada, USA. We classified detailed land‐cover types via object‐based image analysis (OBIA) using Geoeye‐1 at 3‐m resolution (Las Vegas) and QuickBird at 2.4‐m resolution (Phoenix). Spatial autocorrelation (local Morans I) was then used to test for spatial dependence and to determine how clustered or dispersed points were arranged. Next, we used Advanced Spaceborne Thermal Emission and Reflection Radiometer (ASTER) data acquired over Phoenix (daytime on 10 June and nighttime on 17 October 2011) and Las Vegas (daytime on 6 July and nighttime on 27 August 2005) to examine day‐ and nighttime LST with regard to the spatial arrangement of anthropogenic and vegetation features. Local Morans I values of each land‐cover type were spatially correlated to surface temperature. The spatial configuration of grass and trees shows strong negative correlations with LST, implying that clustered vegetation lowers surface temperatures more effectively. In contrast, clustered spatial arrangements of anthropogenic land‐cover types, especially impervious surfaces and open soil, elevate LST. These findings suggest that city planners and managers should, where possible, incorporate clustered grass and trees to disperse unmanaged soil and paved surfaces, and fill open unmanaged soil with vegetation. Our findings are in line with national efforts to augment and strengthen green infrastructure, complete streets, parking management, and transit‐oriented development practices, and reduce sprawling, unwalkable housing development.

Estimation of agricultural soil properties with imaging and laboratory spectroscopy

Abstract Two EO-1 Hyperion images covering a Cicero Creek reservoir of central Indiana were analyzed using partial least squares (PLS) regression to estimate soil properties, including soil moisture, soil organic matter (SOM), total carbon (C), total phosphorus (P), total nitrogen (N), and clay content. PLS results for Hyperion image spectra were compared with those for laboratory measured spectra using several statistics, including the coefficient of determination ( R 2 ) and RPD (the ratio of standard deviation of sample chemical concentration to root mean square error). PLS was conducted in two phases: phase-1 used all samples for calibration to determine outliers and then models were recalibrated after outlier removal; phase-2 split the resulting samples from phase 1 into two subsets for calibration and validation, respectively. Based on R 2 and RPD values, the results from the phase-1 calibration indicate that PLS can estimate all soil properties from laboratory spectra and some soil properties from Hyperion spectra, and the phase 2 results suggest that PLS can predict SOM, total C, and total N using Hyperion reflectance spectra. It was found that spectral resolution has impacts on the PLS performance in estimating the soil properties considered in this investigation.

Characterizing changes in cropping patterns using sequential Landsat imagery: an adaptive threshold approach and application to Phoenix, Arizona

Since the 1970s, the Phoenix Active Management Area has experienced rapid urbanization, mostly through land conversions from agricultural lands to urban land use. Rapid urban expansion and population growth have placed unprecedented pressure on agricultural production in this region. Agricultural intensification, in particular double cropping, has been observed globally as an important response to the growing pressure on land. However, the intensification has a number of negative impacts on water quality, biodiversity, and biogeochemical cycles. Thus, quantifying the spatial pattern of cropping intensity is important for natural resource management. In this study, we developed an adaptive threshold approach to map cropping intensity using time series Landsat data and examined the spatiotemporal patterns of cropping intensity in the Phoenix Active Management Area from 1995 to 2010 at 5-year intervals. To map cropping intensity accurately, the adaptive threshold algorithm was designed specifically to address several issues caused by the complex cropping patterns in the study area. The adaptive threshold method has abilities to (1) distinguish true crop cycles from multiple false phenological peaks, (2) minimize errors caused by data noise and missing data, (3) identify alfalfa and interyear crops and to distinguish alfalfa from double crops, and (4) adapt to temporal profiles with different numbers of observations. The adaptive threshold algorithm is effective in characterizing cropping intensity with overall accuracies exceeding 97%. Results show that there is a dramatic decline in the area of total croplands (46.1%), single crops (46.3%), and double crops (43.4%) during the study period. There was a small conversion (1.9%) from single to double crop from 1995 to 2000, whereas a reverse conversion (1.3%) was observed from 2005 to 2010. Updated and accurate information on the spatial distribution of cropping intensity provide important implications on effective and sustainable cropping practices. In addition, joint investigation on cropping patterns and irrigation water use can shed light on future agricultural water demand, which is of paramount importance in this rapidly expanding arid region.

Partial least squares modeling of Hyperion image spectra for mapping agricultural soil properties

This paper investigated the capacity of Hyperion images coupled with Partial least squares analysis (PLS) for mapping agricultural soil properties. Soil samples were collected from Cicero Creek Watershed of central Indiana, and analyzed for soil moisture content (MC), soil organic matter (SOM), total carbon (C), total phosphorus (P), total nitrogen (N) and clay content. Two scenes of Hyperion images covering the watershed were acquired, calibrated and georeferenced, and image spectra were extracted from them. Two phases of PLS modeling was conducted: all samples were used and outliers were identified and removed in phase 1, and in phase 2, the outlier removed dataset were split into two subsets for calibration and validation. The PLS results for both phases indicate that PLS modeling of Hyperion spectra is effective to predict MC, SOM, total C, and total N, but resulted in low correlations for total P and clay content. The low correlation for total P is attributed to low correlation between SOM and total P. The worst correlation for clay content is due to the low signal-to-noise ratio of Hyperion images in the short wave infrared region. Future work is needed for improving the estimates of total P and clay content.

A Novel Image Classification Algorithm Using Overcomplete Wavelet Transforms

A novel frequency-based classification framework and new wavelet algorithm (Wave-CLASS) is proposed using an over-complete decomposition procedure. This approach omits the downsampling procedure and produces four-texture information with the same dimension of the original image or window at infinite scale. Three image subsets of QuickBird data (i.e., park, commercial, and rural) over a central region in the city of Phoenix were used to examine the effectiveness of the new wavelet over-complete algorithm in comparison with a widely used classical approach (i.e., maximum likelihood). While the maximum-likelihood classifier produced <; 78.29% overall accuracies for all three image subsets, the Wave-CLASS algorithm achieved high overall accuracies-95.05% for the commercial subset (Kappa = 0.94), 93.71% for the park subset (Kappa = 0.93), and 89.33% for the rural subset (Kappa = 0.86). Results from this study demonstrate that the proposed method is effective in identifying detailed urban land cover types in high spatial resolution data.