Mapping Large-Scale Plateau Forest in Sanjiangyuan Using High-Resolution Satellite Imagery and Few-Shot Learning

Monitoring the extent of plateau forests has drawn much attention from governments given the fact that the plateau forests play a key role in global carbon circulation. Despite the recent advances in the remote-sensing applications of satellite imagery over large regions, accurate mapping of plateau forest remains challenging due to limited ground truth information and high uncertainties in their spatial distribution. In this paper, we aim to generate a better segmentation map for plateau forests using high-resolution satellite imagery with limited ground-truth data. We present the first 2 m spatial resolution large-scale plateau forest dataset of Sanjiangyuan National Nature Reserve, including 38,708 plateau forest imagery samples and 1187 handmade accurate plateau forest ground truth masks. We then propose an few-shot learning method for mapping plateau forests. The proposed method is conducted in two stages, including unsupervised feature extraction by leveraging domain knowledge, and model fine-tuning using limited ground truth data. The proposed few-shot learning method reached an F1-score of 84.23%, and outperformed the state-of-the-art object segmentation methods. The result proves the proposed few-shot learning model could help large-scale plateau forest monitoring. The dataset proposed in this paper will soon be available online for the public.

Geophysical mapping of a classical Greek road network: A case study from the city of Elis, Peloponnese

Many ancient Greek cities are characterised by a regular orthogonal road network. These roads are ideal targets for geophysical investigation mainly due to their extensive geographic extent that makes them challenging to define by excavation. Geophysical mapping of these features will contribute to understanding ancient cities as it can provide considerable information about their geographic extent, spatial arrangement and urban dynamics. Large scale multisensor magnetic and electromagnetic induction methods have been used to map the ancient Greek city of Elis in the Peloponnese (Greece). This work complements other investigations that have been undertaken, employing other methods that include the interpretation of high-resolution satellite imagery (Donati and Sarris forthcoming).

Clarifying the Smokescreen of Russian Protected Areas

Although in strictly protected areas no forest management and logging activities should be evident, a preliminary study detected that, even in the 200 areas with the highest protection of Russia, more than 2 Mha of trees have been lost between 2001 and 2018. Nonetheless, a relevant percentage of the actual drivers of tree loss in Russian strictly protected areas was surrounded by uncertainties due to several factors. Here, in an attempt to “clarify the smokescreen of Russian protected areas”, by validating previous remotely sensed data with new high-resolution satellite imagery and aerial images of land-use change, we shed more light on what has happened during the last 20 years. We used the same layer of tree loss from 2001 to 2020 but, instead of intersecting it with the MODIS data that could have been a source of underestimation of burned surfaces, we overlapped it to the layer of tree cover loss by dominant driver. We analysed the main drivers of tree loss in almost 200 strictly protected areas of Russia. We found that although fire is responsible for 75% of the loss in all strictly protected areas, forestry activities still account for 16%, and 9% is due to undefined causes. Therefore, uncontrolled wildfires (including those started before or after logging) and forestry activities are the main causes of 91% of the total tree loss. The combination of wildfires (often started intentionally) and forestry activities (illegally or barely legally put in place) caused a loss of an astonishing 3 million hectares. The fact that ≈10% of Russian tree cover was lost in two decades since 2001 only in strictly protected areas requires high attention by policymakers and important conservation actions to avoid losing other fundamental habitats and species during the next years when climate change and population growth can represent an additional trigger of an already dramatic situation. We call for an urgent response by national and local authorities that should start actively fighting wildfires, arsonists, and loggers even in inhabited remote areas and particularly in those included in strictly protected areas.

Semi-Automatic Extraction and Mapping of Farmlands based on High-Resolution Remote Sensing Images

Extraction of agricultural parcels from high-resolution satellite imagery is an important task in precision agriculture. Here, we present a semi-automatic approach for agricultural parcel detection that achieves high accuracy and efficiency. Unlike the techniques presented in previous literatures, this method is pixel based, and it exploits the properties of a spectral angle mapper (SAM) to develop customized operators to accurately derive the parcels. The main steps of the method are sample selection, textural analysis, spectral homogenization, SAM, thresholding, and region growth. We have systematically evaluated the algorithm proposed on a variety of images from Gaofen-1 wide field of view (GF-1 WFV), Resource 1-02C (ZY1-02C), and Gaofen-2 (GF-2) to aerial image; the accuracies are 99.09% of GF-1 WFV, 84.42% of ZY1-02C, 96.51% and 92.18% of GF-2, and close to 100% of aerial image; these results demonstrated its accuracy and robustness.

CNN-Based Individual Tree Species Classification Using High-Resolution Satellite Imagery and Airborne LiDAR Data

Deep learning (DL) has shown promising performances in various remote sensing applications as a powerful tool. To explore the great potential of DL in improving the accuracy of individual tree species (ITS) classification, four convolutional neural network models (ResNet-18, ResNet-34, ResNet-50, and DenseNet-40) were employed to classify four tree species using the combined high-resolution satellite imagery and airborne LiDAR data. A total of 1503 samples of four tree species, including maple, pine, locust, and spruce, were used in the experiments. When both WorldView-2 and airborne LiDAR data were used, the overall accuracies (OA) obtained by ResNet-18, ResNet-34, ResNet-50, and DenseNet-40 were 90.9%, 89.1%, 89.1%, and 86.9%, respectively. The OA of ResNet-18 was increased by 4.0% and 1.8% compared with random forest (86.7%) and support vector machine (89.1%), respectively. The experimental results demonstrated that the size of input images impacted on the classification accuracy of ResNet-18. It is suggested that the input size of ResNet models can be determined according to the maximum size of all tree crown sample images. The use of LiDAR intensity image was helpful in improving the accuracies of ITS classification and atmospheric correction is unnecessary when both pansharpened WorldView-2 images and airborne LiDAR data were used.

Dispersal and fire limit Arctic shrub expansion

Arctic shrub expansion has been widely reported in recent decades, with large impacts on carbon budgets, albedo, and warming rates in high latitudes. However, predicting shrub expansion across regions remains challenging because the underlying controls remain unclear. Observational studies and models typically use relationships between observed shrub presence and current environmental suitability (climate and topography) to predict shrub expansion, but such approaches omit potentially important biotic-abiotic interactions and non-stationary relationships. Here, we use long-term high-resolution satellite imagery across Alaska and western Canada to show that observed shrub expansion has not been controlled by environmental suitability during 1984-2014, but rather can only be explained by accounting for seed dispersal and fire. These findings provide the impetus for better observations of recruitment and for incorporating currently underrepresented processes of seed dispersal and fire in land models to project shrub expansion and future climate feedbacks. Integrating these dynamic processes with projected fire extent and climate, we estimate that shrubs will expand into 25% of the non-shrub tundra by 2100, in contrast to 39% predicted using a relationship with increasing suitability alone. Thus, using environmental suitability alone likely overestimates and misrepresents the spatial pattern of shrub expansion and its associated carbon sink.

URBAN FLOOD IMPACTS, FLOOD WATER QUALITY AND RISK MAPPING OF OLODO AREA, IBADAN, NIGERIA

This study assessed urban flood impact, flood water quality and vulnerability around Olodo area of Ibadan region, Nigeria. The study employed remote sensing and GIS techniques in creating vulnerability and risk maps. Digital terrain model (DTM) was used to get the topography of the study area. Footprints of buildings along the Egberi riverbank and flood plain in Olodo were created in the GIS environment from high resolution satellite imagery. Buffering operation was conducted to classify the buildings into risk zones based on closeness to the riverbank using ArcGIS 10.0. The study revealed that 326 buildings were within the very vulnerable and vulnerable zones because they were less than 15.2m away from the riverbank. The characteristics of water quality change during the flood and non-flood periods. TSS, DO, NOD, and COD were all higher during the flood event. Microbial analysis showed that water quality levels in the floodwater exceeded water quality standards (e.g., the coliform excess from 10 to 10,000 times), and thus this may be a health risk for local people during flood events. Concentration of Escherichia coli (E. coli) ranged from 484 to 1290 cfu/100 mL during flooding compared to 192 to 295 cfu/100 mL after flood. Salmonella was found to be high ranging from 659 to 1840 cfu/100 mL during flooding compared to 530 to 1034 cfu/100 mL after flooding.

Monitoring of Land Use Suitability Based on Detail Urban Spatial Plan Data 2018-2038 (Case study: Gayungan District, Surabaya City)

Land use planning in an area will refer to the regulations that have been established by the City Planning Office of each region. This is because each region has the authority to plan spatial plans in their respective regions. The border area is an area whose land use can be influenced by two different regional regulations. This is because the border area is a special area located on the border between two regions, each of which has spatial planning regulations. Gayungan District is one of the sub-districts included in the Border Area between Surabaya City and Sidoarjo Regency. To prevent overlapping spatial regulations that may occur in Gayungan District, it is necessary to monitor the suitability of land use using geographic information system technology and remote sensing. The data used in this study are the 2019 Gayungan District RDTRK Map and very high resolution satellite imagery of Pleiades Surabaya City 2019. The method used is the Object Based Image Analysis (OBIA) method. The result of this research is the suitability of land use in Gayungan District. A land use can be said to be suitable if the existing land use is in accordance with the land use in the plan. Meanwhile, land use is said to be inappropriate if the existing land use is different from the planned spatial use. All land use classes in Gayungan District in 2019 had a higher percentage of unsuitable land than the percentage of suitable land. Each percentage of land is not suitable for each land use class, namely: water body class by 92.593%, road class by 78.035%, industrial class by 77.838%, defense class by 76.706%, green open space class by 69.736%, and residential class by 52,27%. So it can be said that the land use in Gayungan District in 2019 was not in accordance with the plans in the City Spatial Detail Plan Map for 2018-2038, but the land use could be appropriate in its designation for the future, because there is a possibility of development for residential class, industrial class, and defense class on open land that is still widely available.

The agglomeration and dispersion dichotomy of human settlements on Earth

Human settlements on Earth are scattered in a multitude of shapes, sizes and spatial arrangements. These patterns are often not random but a result of complex geographical, cultural, economic and historical processes that have profound human and ecological impacts. However, little is known about the global distribution of these patterns and the spatial forces that creates them. This study analyses human settlements from high-resolution satellite imagery and provides a global classification of spatial patterns. We find two emerging classes, namely agglomeration and dispersion. In the former, settlements are fewer than expected based on the predictions of scaling theory, while an unexpectedly high number of settlements characterizes the latter. To explain the observed spatial patterns, we propose a model that combines two agglomeration forces and simulates human settlements’ historical growth. Our results show that our model accurately matches the observed global classification (F1: 0.73), helps to understand and estimate the growth of human settlements and, in turn, the distribution and physical dynamics of all human settlements on Earth, from small villages to cities.