Satellite-derived Land Cover Research Articles

Remapping California's wildland urban interface: A property-level time-space framework, 2000–2020

Wildland fire poses a significant risk to property and people, and these risks tend to be highest in places where flammable, wildland vegetation abuts urban development. Traditionally, efforts to map these areas of intersection, known as the Wildland Urban Interface (WUI), have relied on satellite-derived land cover and census data. While these types of data are valuable for broadly delineating WUI areas, they are constrained in their ability to specifically identify at-risk residential properties and their characteristics over time. This article addresses these challenges through a multi-time-step, property-level framework that enhances current approaches based on moving window mapping, the Time Step Moving Window (TSMW) framework. We implement this framework on approximately 10 million geo-located property records from Zillow's Transaction and Assessment Database for California. Using this framework, we generate estimates of California's WUI from 2000 to 2020 and find that our time stepped building scale maps have allowed us to classify residential homes at-risk to burns with greater accuracy compared to time stepped Census polygon based maps. Based on these maps, we further estimate that approximately $1.34 Trillion, or 40%, of California's improved residential property value is situated within the WUI. Our approach provides a high-precision strategy for delineating at-risk residences and calculating localized financial costs associated with future hazard events.

Possible Integration of Soil Information into Land Degradation Analysis for the United Nations (UN) Land Degradation Neutrality (LDN) Concept: A Case Study of the Contiguous United States of America (USA)

Soil makes important contributions to the United Nations (UN) Land Degradation Neutrality (LDN) concept and targets; however, currently, soil is not integrated into measurable information (e.g., indicators, metrics) to monitor land degradation (LD) patterns and trends. This study examines the role of soil in LDN in the UN Convention to Combat Desertification (UNCCD), and UN Sustainable Development Goal (SDG 15: Life on Land). This study is specifically focused on the LDN and biodiversity loss as they relate to an indicator 15.3.1 Proportion of land that is degraded over total land area. Tracking of LD status can be improved by using detailed soils databases combined with satellite-derived land cover maps. This study has applied these newly improved methods to quantify and map the anthropogenic LD status and trends in the contiguous United States of America (USA), as well as to identify potential land areas for nature-based solutions (NBS) to compensate for LD. Anthropogenic LD in 2016 in the contiguous USA affected over two million square kilometers, about one-third of the country’s total area, with high variability by state. Between 2001 and 2016, LD in the USA showed an overall increase of 1.5%, with some states exhibiting increases in degraded land while other states had overall improvements to their land. All ten soil orders present in the contiguous USA have been anthropogenically degraded, with Mollisols, Alfisols, and Vertisols having the highest LD levels. Compensating for LD requires a variety of strategies and measures (e.g., NBS), which often require additional land. In 2016, the potential land area for NBS was over two million square kilometers, an area approximately equal to that of degraded land. Some of the states that have high proportions of land available for potential NBS are dominated by soils (Aridisols) typical of deserts and therefore may have less promise for NBS. The variability of LD needs to be evaluated at finer spatial scales for realistic LDN analysis.

Using individual approach to examine the association between urban heat island and preterm birth: A nationwide cohort study in China

BackgroundEvidence suggests that maternal exposure to heat might increase the risk of preterm birth (PTB), but no study has investigated the effect from urban heat island (UHI) at individual level. AimsOur study aimed to investigate the association between individual UHI exposure and PTB. MethodsWe utilized data from the ongoing China Birth Cohort Study (CBCS), encompassing 103,040 birth records up to December 2020. UHI exposure was estimated for each participant using a novel individual assessment method based on temperature data and satellite-derived land cover data. We used generalized linear mixed-effects models to estimate the association between UHI exposure and PTB, adjusting for potential confounders including maternal characteristics and environmental factors. ResultsConsistent and statistically significant associations between UHI exposure and PTB were observed up to 21 days before birth. A 5 °C increment in UHI exposure was associated with 27 % higher risk (OR = 1.27, 95 % confident interval: 1.20, 1.34) of preterm birth in lagged day 1. Stratified analysis indicated that the associations were more pronounced in participants who were older, had higher pre-pregnancy body mass index level, of higher socioeconomic status and living in greener areas. ConclusionMaternal exposure to UHI was associated with increased risk of PTB. These findings have implications for developing targeted interventions for susceptible subgroups of pregnant women. More research is needed to validate our findings of increased risk of preterm birth due to UHI exposure among pregnant women.

Impact of Satellite-Derived Land Cover Resolution Using Machine Learning and Hydrological Simulations

This study carefully assesses the capability of supervised machine learning classification algorithms in identifying land cover (LC) in the context of the Jhelum River basin in Kashmir. Sentinel 2 and Landsat 8 high-resolution data from two satellite sources were used. Through preprocessing techniques, we removed any potential noise inherent to satellite imagery and assured data consistency. The study then utilized and compared the skills of the supervised algorithms random forest (RF) and support vector machine (SVM). A hybrid approach, amalgamating classifications from both methods, was also tested for potential synergistic enhancements in accuracy. Using a stratified random sampling approach for validation, the SVM algorithm emerged with a commendable accuracy rate of 82.5%. Using simulations from 2000 to 2015, the soil and water assessment tool (SWAT) model was used to further explore the hydrological effects of LC alterations. Between 2009 and 2019, there were discernible changes in the land cover, with a greater emphasis on ranges, forests, and agricultural plains. When these changes were combined with the results of the hydrologic simulation, a resultant fall in average annual runoff—from above 700 mm to below 600 mm—was seen. With runoff values possibly ranging between 547 mm and 747 mm, the statistics emphasize the direct effects of urban communities encroaching upon forest, agricultural, and barren lands. This study concludes by highlighting the crucial role that technical pipelines play in enhancing LC classifications and by providing suggestions for future water resource estimation and hydrological impact evaluations.

A geospatial assessment of soil properties to identify the potential for crop rotation in rice systems

Crop rotation is one strategy for adapting agroecosystems to a framework that balances ecological diversity, sustainability, and food production. The Sacramento Valley, one of the most productive rice growing regions in the US, faces sustainability challenges including increasing herbicide resistant weed pressure and water use restrictions. Increasing crop diversity may help address these challenges, but this region has unique soil attributes including high clay content, salinity, alkalinity, and cemented subsurface layers, and the degree to which these soil properties influence crop rotation decisions remain unclear. The objectives of this study were to quantify the extent of crop rotation in this region, compare soil properties for rotated and continuous rice fields, and assess the potential for expanding rotations based on the geographic coverage of influential soil variables. Using satellite derived land cover data for 2007–2021, our analysis shows that only ∼5000 ha are in rotation with rice, while 220,000 ha are in continuous rice production. This land cover information is fused with SSURGO soil maps in a spatial random forest model. The modeling approach indicates that fields with soil pH between 6.5 and 8, EC between 0.5 and 2 (ds m-1), and saturated hydraulic conductivity less than 2 (μm s-1) are more likely to be rotated. However, we estimate that only 11% of the continuous rice area has all three of these soil properties combined, suggesting soil limitations are an important constraint. This research highlights a method for evaluating land use decisions in relation to spatial variability of soil properties to better understand barriers to agroecological diversification.

Comparing Global Sentinel-2 Land Cover Maps for Regional Species Distribution Modeling

Mapping the spatial and temporal dynamics of species distributions is necessary for biodiversity conservation land-use planning decisions. Recent advances in remote sensing and machine learning have allowed for high-resolution species distribution modeling that can inform landscape-level decision-making. Here we compare the performance of three popular Sentinel-2 (10-m) land cover maps, including dynamic world (DW), European land cover (ELC10), and world cover (WC), in predicting wild bee species richness over southern Norway. The proportion of grassland habitat within 250 m (derived from the land cover maps), along with temperature and distance to sandy soils, were used as predictors in both Bayesian regularized neural network and random forest models. Models using grassland habitat from DW performed best (RMSE = 2.8 ± 0.03; average ± standard deviation across models), followed by ELC10 (RMSE = 2.85 ± 0.03) and WC (RMSE = 2.87 ± 0.02). All satellite-derived maps outperformed a manually mapped Norwegian land cover dataset called AR5 (RMSE = 3.02 ± 0.02). When validating the model predictions of bee species richness against citizen science data on solitary bee occurrences using generalized linear models, we found that ELC10 performed best (AIC = 2278 ± 4), followed by WC (AIC = 2367 ± 3), and DW (AIC = 2376 ± 3). While the differences in RMSE we observed between models were small, they may be significant when such models are used to prioritize grassland patches within a landscape for conservation subsidies or management policies. Partial dependencies in our models showed that increasing the proportion of grassland habitat is positively associated with wild bee species richness, thereby justifying bee conservation schemes that aim to enhance semi-natural grassland habitat. Our results confirm the utility of satellite-derived land cover maps in supporting high-resolution species distribution modeling and suggest there is scope to monitor changes in species distributions over time given the dense time series provided by products such as DW.

Landslides Detection and Mapping with an Advanced Multi-Temporal Satellite Optical Technique

Landslides are catastrophic natural phenomena occurring as a consequence of climatic, tectonic, and human activities, sometimes combined among them. Mostly due to climate change effects, the frequency of occurrence of these events has quickly grown in recent years, with a consequent increase in related damage, both in terms of loss of human life and effects on the involved infrastructures. Therefore, implementing properly actions to mitigate consequences from slope instability is fundamental to reduce their impact on society. Satellite systems, thanks to the advantages offered by their global view and sampling repetition capability, have proven to be valid tools to be used for these activities in addition to traditional techniques based on in situ measurements. In this work, we propose an advanced multitemporal technique aimed at identifying and mapping landslides using satellite-derived land cover information. Data acquired by the Multispectral Instrument (MSI) sensor aboard the Copernicus Sentinel-2 platforms were used to investigate a landslide affecting Pomarico city (southern Italy) in January 2019. Results achieved indicate the capability of the proposed methodology in identifying, with a good trade-off between reliability and sensitivity, the area affected by the landslide not just immediately after the event, but also a few months later. The technique was implemented within the Google Earth Engine Platform, so that it is completely automatic and could be applied everywhere. Therefore, its potential for supporting mitigation activities of landslide risks is evident.

Methodology for Lidar Monitoring of Biomass Burning Smoke in Connection with the Land Cover

Lidar measurements of 11 smoke layers recorded at Măgurele, Romania, in 2014, 2016, and 2017 are analyzed in conjunction with the vegetation type of the burned biomass area. For the identified aerosol pollution layers, the mean optical properties and the intensive parameters in the layers are computed. The origination of the smoke is estimated by the means of the HYSPLIT dispersion model, taking into account the location of the fires and the injection height for each fire. Consequently, for each fire location, the associated land cover type is acquired by satellite-derived land cover products. We explore the relationship between the measured intensive parameters of the smoke layers and the respective land cover of the burned area. The vegetation type for the cases we analyzed was either broadleaf crops or grasses/cereals. Overall, the intensive parameters are similar for the two types, which can be associated with the fact that both types belong to the broader group of agricultural crops. For the cases analyzed, the smoke travel time corresponding to the effective predominant vegetation type is up to 2.4 days.

Assessment of land use change in the Thuma forest reserve region of Malawi, Africa

A study was conducted in Thuma area in central Malawi to quantify contemporary land cover and to explore the degree of land use change in the Thuma forest reserve area of Malawi by analysing and comparing satellite-derived land cover maps from 1997, 2007 and 2017. The study was carried out using Remote Sensing and Geographic Information System (GIS), focusing on analysis of Landsat 5 ETM and Landsat 8 ORI/TIRS satellite images. The classification was conducted for the following distinct classes; closed forest, open forest, shrubland, savanna grassland, agriculture fields, and water. The analysis revealed that closed forest diminished from 19% in 1997 to 10% in 2007 to 6% in 2017. Open forest reduced from 30% to 21% from 1997 to 2007 but increased to 22% in 2017. Agriculture area almost doubled from 37% in 1997 to 64% in 2017. The actual area from 1997 to 2017, shows that closed forest has reduced from 7,000 ha to 3,000 ha while open forest from 12,900 ha to 7800 ha. Savanna grassland has doubled from 5,900 ha to 13,000 ha. However, future studies should use modern satellites such as Sentinel and Landsat 9 for improved quantification of changes. The findings show that even the protected forest reserve (previously dominated by closed forest) is not fully protected from deforestation by local communities. Government and other stakeholders should devise measures to meet the needs of the surrounding communities and the ecological/biophysical needs of the reserves. Based on this study, issues of re-demarcation of the forest reserve and the accessed area should also be explored. This study serves as a reference for the management of Thuma Forest Reserve as a refuge for natural tree species, rivers that harbour endemic fish species (Opsaridium microlepis and Opsaridium microcephalis) and the sustainable management of endangered elephants in the reserve.

Habitat selection and food choice of White-naped Cranes (Grus vipio) at stopover sites based on satellite tracking and stable isotope analysis

By combining satellite tracking, land-cover extracted from Landsite 8 images, and the traditional stable isotope analysis, we studied the habitat selection and food preference of a vulnerable migratory waterbird, the White-naped Crane (Grus vipio), in one of its key stagging sites, the Shandian River Basin in the semi-arid northern China, to provide knowledge that is critical for its conservation in the Anthropocene. Our results showed that the White-naped Cranes used both uplands and natural wetlands in the stopover site. While the cranes used farmland and natural land cover equally as night-time roosting grounds, they spent most daytime foraging at farmlands. Despite the extensive usage of croplands as their foraging ground, the Bayesian mixing models based on stable isotopic analysis revealed that crop residues after harvesting, such as Maize (Zea mays) and Naked Oat (Avena chinensis), were only a small fraction of the White-naped Cranes’ diet (∼19%), and their diet composited mainly natural plants, such as Allium ledebourianum, Potentilla anserina, and P. tanacetifoli. Moreover, more than 20% of the total wetlands in the region were modelled as home range of the cranes. On contrast, less than 10% of croplands and about 1% of the unused uplands were identified as home range. In addition, the entire core habitats were located in natural wetlands. Our findings demonstrated the importance of natural wetlands for the survival of the threatened crane. However, the satellite-derived land cover data showed that croplands increased rapidly in the last decade in this area, at the expense of natural wetlands. With the sharp decrease of White-naped Crane population in China, the conservation of stopover sites becomes imperative. Based on our analysis, we recommend the following management actions: conserving adequate natural wetland area, regulating anthropogenic pressures such as the use of herbicides, expanding the duration and extent of current conservation regulations, establishing a comprehensive monitoring program, and initiating basin-scale ecological restoration, for effective conservation of this threatened species. These integrated conservation strategies for migratory waterbirds are necessary, considering the rapid land-cover changes and agricultural expansion that have been occurring in East Asian-Australasian Flyway, especially in the semi-arid temperate zone.

Revisiting the use of red and near-infrared reflectances in vegetation studies and numerical climate models

Surface reflectance data acquired in red and near-infrared spectra by remote sensing sensors are traditionally applied to construct various vegetation indices (VIs), which are related to vegetation biophysical parameters. Most VIs use pre-defined weights (usually equal to 1) for the red and NIR reflectance values, therefore constraining particular weights for red and NIR during the VI design phase, and potentially limiting capabilities of the VI to explain an independent variable. In this paper, we propose an approach to estimate biophysical variables, such as Leaf Area Index (LAI), Canopy Chlorophyll Content (CCC) and Fraction of Photosynthetically Active Radiation (FPAR) absorbed by green vegetation, represented as linear combinations of the red and NIR reflectances with weights determined empirically from observations and radiative transfer model (PROSAIL) simulations. The proof of concept is first tested on available close-range observations over maize and soybean crops in Nebraska, USA. The empirical results compare well with those from PROSAIL model simulations. The proposed LAI model is then used with data from Landsat 8, Sentinel-2 and Planet/Dove, and the results are validated with in situ LAI measurements in Ukraine. We show that the weights on red and NIR reflectances are vegetation-specific and stable in time. The approach is further tested on crops and forests in the conterminous USA and on a global scale using MODIS LAI and FPAR products as proxies for “ground observations”. These LAI and FPAR, however, are not independently measured but derived from the corresponding remotely sensed reflectances, which precludes recommending a final set of the weights/coefficients for the users, and, thus, should be considered mostly for demonstrating the concept. The results for crop types, other than maize and soybean, and for all forests are conceptual and need to be tested with real ground data. It was, however, encouraging to see that the derived maps of coefficients/weights exhibit regular patterns over the globe compatible with those of vegetation classes and crop types. Tedious and thorough work on compiling available in situ measurements on various crops and forests needs to be accomplished prior to large-scale applications, and the method needs to be further tested and proven that it works at a large scale.The proposed parameterization may be attractive for global studies of various sub-classes of vegetation, once the parameter coefficients are established, validated, tabulated and their stability verified. Ultimately, this approach may provide quantification of vegetation traits for the past decades and be a useful asset for climate models that include satellite-derived land cover classifications and vegetation variables for simulating surface fluxes.This is a conceptual paper, with a proof-of-concept supported by observations over two crops, for which we had close-range observations. It is not a technical note, which would provide users with a recommended set of coefficients for global applications. Our intent was to develop a paradigm, which could ultimately be useful in global models.

The land–energy–water nexus of global bioenergy potentials from abandoned cropland

Bioenergy is a key option in climate change mitigation scenarios. Growing perennial grasses on recently abandoned cropland is a near-term strategy for gradual bioenergy deployment with reduced risks for food security and the environment. However, the extent of global abandoned cropland, bioenergy potentials and management requirements are unclear. Here we integrate satellite-derived land cover maps with a yield model to investigate the land–energy–water nexus of global bioenergy potentials. We identified 83 million hectares of abandoned cropland between 1992 and 2015, corresponding to 5% of today’s cropland area. Bioenergy potentials are 6–39 exajoules per year (11–68% of today’s bioenergy demand), depending on multiple local and management factors. About 20 exajoules per year can be achieved by increasing today’s global cropland area and water use by 3% and 8%, respectively, and without production inside biodiversity hotspots or irrigation in water-scarce areas. The consideration of context-specific practices and multiple environmental dimensions can mitigate trade-offs of bioenergy deployment. Bioenergy from grasses is a key option to mitigate climate change. This study finds that recently abandoned cropland could help meet 11–68% of today’s bioenergy demand.

Empirical Bayes small area prediction under a zero-inflated lognormal model with correlated random area effects.

Many variables of interest in agricultural or economical surveys have skewed distributions and can equal zero. Our data are measures of sheet and rill erosion called Revised Universal Soil Loss Equation - 2 (RUSLE2). Small area estimates of mean RUSLE2 erosion are of interest. We use a zero-inflated lognormal mixed effects model for small area estimation. The model combines a unit-level lognormal model for the positive RUSLE2 responses with a unit-level logistic mixed effects model for the binary indicator that the response is nonzero. In the Conservation Effects Assessment Project (CEAP) data, counties with a higher probability of nonzero responses also tend to have a higher mean among the positive RUSLE2 values. We capture this property of the data through an assumption that the pair of random effects for a county are correlated. We develop empirical Bayes (EB) small area predictors and a bootstrap estimator of the mean squared error (MSE). In simulations, the proposed predictor is superior to simpler alternatives. We then apply the method to construct EB predictors of mean RUSLE2 erosion for South Dakota counties. To obtain auxiliary variables for the population of cropland in South Dakota, we integrate a satellite-derived land cover map with a geographic database of soil properties. We provide an R Shiny application called viscover (available at https://lyux.shinyapps.io/viscover/) to visualize the overlay operations required to construct the covariates. On the basis of bootstrap estimates of the mean square error, we conclude that the EB predictors of mean RUSLE2 erosion are superior to direct estimators.

Observed earlier start of the growing season from middle to high latitudes across the Northern Hemisphere snow-covered landmass for the period 2001–2014

Vegetation phenology in spring has received much attention for its importance to terrestrial ecosystem carbon exchange and climate–biosphere interactions studies. Through control on surface water and heat balance, snow cover largely impacts on spring vegetation phenology. However, under the background of global warming and rapid reduction of spring snow cover extent across the Northern Hemisphere (NH), the responses of spring vegetation phenology have not been well documented. Using two satellite-derived land cover dynamic datasets and 420 in situ vegetation phenology observations from five filed datasets, this study evaluated the accuracy of satellite-derived vegetation phenology datasets and explored the changes of start of the growing season (SOS) across the NH snow-covered landmass for the period 2001–2014. Compared with MEaSUREs VIPPHEN, the MODIS SOS maps displayed higher accuracy in capture the real SOS climatology by validating with in situ observations (R2 = 0.67, bias = −3.99 d). Moreover, evidences from MODIS SOS maps pointed out that the SOS advanced by approximately 2.36 d in NH middle to high latitudes (43.5°N–70.0°N), but delayed by about 0.53 d in lower latitudes (33.0°N–43.5°N) from 2001 to 2014. The contrast SOS anomalies across the NH snow-covered landmass were further proved by changes in spring NDVI derived from GIMMS in the corresponding period. In addition, the observed changes in SOS were consistent with the spatiotemporal pattern of spring snow end date found in previous studies, indicating vegetation phenology changes should be taken into account in estimating the impacts of snow in climate–biosphere interactions studies.

Characterizing uncertainties in the ESA-CCI land cover map of the epoch 2010 and their impacts on MPI-ESM climate simulations

Limitations of mapping land surface properties and their conversion into climate model boundary conditions are major sources of uncertainty in climate simulations. In this paper, the range of the largest possible uncertainty in satellite-derived land cover (LC) map is estimated and its impact on climate simulations is quantified with the Earth System Model of the Max-Planck Institute for Meteorology utilizing prescribed sea surface temperature and sea ice. Two types of uncertainty in the LC map are addressed: (i) uncertainty due to classification algorithm of spectral reflectance into LC classes, and (ii) uncertainty due to conversion of LC classes into the climate model vegetation distribution. For forest cover, each of them is about the same order of magnitude as the uncertainty range in recent observations (∼± 700 Mha). Superposing two sources of uncertainty results in LC maps that feature the range of vegetation deviation that is about the same order of magnitude as the recent (since year 1700) forest loss due to agriculture (forest cover uncertainty range ∼± 1700 Mha). These uncertainties in vegetation distribution lead to noticeable variations in near-surface climate variables, local, regional, and global climate forcing. Temperature does not show significant uncertainty in global mean, but rather exhibits regional deviations with an opposite response to LC uncertainty that compensate each other in the global mean (e.g., albedo feedback controls temperature in boreal North America resulting in cooling (warming) with decrease (increase) of vegetation while evaporative cooling controls temperature in South America and sub-Saharan Africa resulting in cooling (warming) with increase (decrease) of vegetation). Large-scale circulation is also affected by the LC uncertainty, and consequently precipitation pattern as well. It is demonstrated that precipitation uncertainty in the monsoonal regions are about the same order of magnitude as in previous studies with idealized perturbations of vegetation. These findings indicate that the range of uncertainty in satellite-derived vegetation maps for climate models is about the same order of magnitude as the uncertainty in recent observations of forest cover or as the forest lost due to agriculture. Consequently, climate simulations have a similar range of uncertainty in variables representing near-surface climate as the observed climate change due to land use. Hence, more accurate methods are needed for mapping and converting LC properties into model vegetation in order to increase reliability of climate model simulations.

Urbanization as a risk factor for aortic stiffness in a cohort in India.

Urbanization is associated with higher prevalence of cardiovascular disease worldwide. Aortic stiffness, as measured by carotid-femoral pulse wave velocity is a validated predictor of cardiovascular disease. Our objective was to determine the association between urbanization and carotid-femoral pulse wave velocity. The analysis included 6166 participants enrolled in an ongoing population-based study (mean age 42 years; 58% female) who live in an 80 × 80 km region of southern India. Multiple measures of urbanization were used and compared: 1) census designations, 2) satellite derived land cover (crops, grass, shrubs or trees as rural; built-up areas as urban), and 3) distance categories based on proximity to an urban center. The association between urbanization and carotid-femoral pulse wave velocity was tested in sex-stratified linear regression models. People residing in urban areas had significantly (p < 0.05) elevated mean carotid-femoral pulse wave velocity compared to non-urban populations after adjustment for other risk factors. There was also an inverse association between distance from the urban center and mean carotid-femoral pulse wave velocity: each 10 km increase in distance was associated with a decrease in mean carotid-femoral pulse wave velocity of 0.07 m/s (95% CI: -0.09, -0.06 m/s). The association was stronger among older participants, among smokers, and among those with other cardiovascular risk factors. Further research is needed to determine which components in the urban environment are associated with higher carotid-femoral pulse wave velocity.

Modelling and mapping tick dynamics using volunteered observations

BackgroundTick populations and tick-borne infections have steadily increased since the mid-1990s posing an ever-increasing risk to public health. Yet, modelling tick dynamics remains challenging because of the lack of data and knowledge on this complex phenomenon. Here we present an approach to model and map tick dynamics using volunteered data. This approach is illustrated with 9 years of data collected by a group of trained volunteers who sampled active questing ticks (AQT) on a monthly basis and for 15 locations in the Netherlands. We aimed at finding the main environmental drivers of AQT at multiple time-scales, and to devise daily AQT maps at the national level for 2014.MethodTick dynamics is a complex ecological problem driven by biotic (e.g. pathogens, wildlife, humans) and abiotic (e.g. weather, landscape) factors. We enriched the volunteered AQT collection with six types of weather variables (aggregated at 11 temporal scales), three types of satellite-derived vegetation indices, land cover, and mast years. Then, we applied a feature engineering process to derive a set of 101 features to characterize the conditions that yielded a particular count of AQT on a date and location. To devise models predicting the AQT, we use a time-aware Random Forest regression method, which is suitable to find non-linear relationships in complex ecological problems, and provides an estimation of the most important features to predict the AQT.ResultsWe trained a model capable of fitting AQT with reduced statistical metrics. The multi-temporal study on the feature importance indicates that variables linked to water levels in the atmosphere (i.e. evapotranspiration, relative humidity) consistently showed a higher explanatory power than previous works using temperature. As a product of this study, we are able of mapping daily tick dynamics at the national level.ConclusionsThis study paves the way towards the design of new applications in the fields of environmental research, nature management, and public health. It also illustrates how Citizen Science initiatives produce geospatial data collections that can support scientific analysis, thus enabling the monitoring of complex environmental phenomena.

R.pi: A grass gis package for semi‐automatic spatial pattern analysis of remotely sensed land cover data

Abstract Analysing the changing spatial patterns of landscapes due to climate change or anthropogenic impact is important for various disciplines. Land cover change and its resulting modification of spatial patterns in the landscape influence various geographical or ecological parameters. Changing formerly continuous into discontinuous ecosystems due to land cover conversion causes isolated fragments in the landscape. Maintaining the connectivity of a fragmented landscape is relevant for, e.g. in nutrient cycle, water‐runoff or species population persistence. Satellite imagery derived land cover can be used to analyse continuously the changing spatial arrangement of land cover types. However, analyses are computer intensive and require robust and efficient processing routines. We developed a patch‐based spatial analysis system (r.pi) integrated natively into a Free and Open Source GIS (grass gis) to be able to analyse large amounts of satellite derived land cover data in a semi‐automatic manner, and to ensure high reproducibility and robustness. Various established and newly developed indices for spatial pattern analysis are provided in this program, to derive further meaningful information like spatial configuration, patch irreplaceability or connectivity of fragments based on a dispersal model approach.

Landsat-based classification in the cloud: An opportunity for a paradigm shift in land cover monitoring

Satellite-derived land cover maps play an important role in many applications, including monitoring of smallholder-dominated agricultural landscapes. New cloud-based computing platforms and satellite sensors offer opportunities for generating land cover maps designed to meet the spatial and temporal requirements of specific applications. Such maps can be a significant improvement compared to existing products, which tend to be coarser than 300m, are often not representative of areas with fast-paced land use change, and have a fixed set of cover classes. Here, we present two approaches for land cover classification using the Landsat archive within Google Earth Engine. Random forest classification was performed with (1) season-based composites, where median values of individual bands and vegetation indices were generated from four years for each of four seasons, and (2) metric-based composites, where different quantiles were computed for the entire four-year period. These approaches were tested for six land cover types spanning over 18,000 locations in Zambia, with ground “truth” determined by visual inspection of high-resolution imagery from Google Earth. The methods were trained on 30% of these points and tested on the remaining 70%, and results were also compared with existing land cover products. Overall accuracies of about 89% were achieved for the season- and metric-based approaches for individual classes, with 93% and 94% accuracy for distinguishing cropland from non-cropland. For the latter task, the existing Globeland30 dataset based on Landsat had much lower accuracies (around 77% on average), as did existing cover maps at coarser resolutions. Overall, the results support the use of either season or metric-based classification approaches. Both produce better results than those obtained from previous classifiers, which supports a general paradigm shift away from dependence on standard static products and towards custom generation of on-demand cover maps designed to fulfill the needs of each specific application.

Air quality and its response to satellite-derived urban form in the Yangtze River Delta, China

Air pollution is one of the top environmental concerns and causes of deaths and various diseases worldwide. An important question for sustainable development is to what extent urban design can improve or degrade urban air quality. In this article, we explored the relationship between ground-based observations of air pollution and urban form in the Yangtze River Delta (YRD), the largest metropolitan region in China. We analyzed six criteria pollutants (SO2, NO2, PM10, PM2.5, CO, O3) and summarized metric (air quality index, AQI) from 129 ambient air quality monitoring stations during 2015. Urban form was characterized using six spatial metrics, including the size, shape, regularity, fragmentation and traffic coupling factor of urban patches, based on satellite-derived land cover data. The results indicated that: (1) PM2.5, PM10 and O3 were three primary pollutants in the YRD. The annual average AQI was 79, and the air quality was “moderate” for human health, with the highest and lowest AQI appeared in winter (107) and summer (60). Moreover, the air quality of the southern areas (Zhejiang province, AQI: 68) was generally better than the northern parts (Jiangsu province, AQI: 86). (2) Through the size and shape of urban patches, urban form had a significant effect on urban air quality in the YRD. PARA_MN (Mean Perimeter-area ratio), ENN_MN (Mean Euclidean Nearest Neighbor Distance), CA (Total Urban Area) and NP (Number of urban patches) had the most significant impacts on air quality. PM10 and PM2.5 were two important pollutants highly positively related to CA and NP, while negatively related to PARA_MN and ENN_MN. In addition, the polycentric urban form was associated with high air quality. (3) Land use configuration was an important indicator to describe the urban air quality. When buffer distance of spatial scale was 25km, air quality showed the highest correlation with forest coverage. A high forest coverage rate contributed to the better air quality, increasing or preserving the forested areas would help mitigate the air pollution.