Modeling flood susceptibility zones using hybrid machine learning models of an agricultural dominant landscape of India.

Flooding events are determining a significant amount of damages, in terms of economic loss and also casualties in Asia and Pacific areas. Due to complexity and ferocity of severe flooding, predicting flood-prone areas is a difficult task. Thus, creating flood susceptibility maps at local level is though challenging but an inevitable task. In order to implement a flood management plan for the Balrampur district, an agricultural dominant landscape of India, and strengthen its resilience, flood susceptibility modeling and mapping are carried out. In the present study, three hybrid machine learning (ML) models, namely, fuzzy-ANN (artificial neural network), fuzzy-RBF (radial basis function), and fuzzy-SVM (support vector machine) with 12 topographic, hydrological, and other flood influencing factors were used to determine flood-susceptible zones. To ascertain the relationship between the occurrences and flood influencing factors, correlation attribute evaluation (CAE) and multicollinearity diagnostic tests were used. The predictive power of these models was validated and compared using a variety of statistical techniques, including Wilcoxon signed-rank, t-paired tests and receiver operating characteristic (ROC) curves. Results show that fuzzy-RBF model outperformed other hybrid ML models for modeling flood susceptibility, followed by fuzzy-ANN and fuzzy-SVM. Overall, these models have shown promise in identifying flood-prone areas in the basin and other basins around the world. The outcomes of the work would benefit policymakers and government bodies to capture the flood-affected areas for necessary planning, action, and implementation.

Study of habitat quality assessment using geospatial techniques in Keoladeo National Park, India.

Wetlands are one of the most productive ecosystems on the Earth. They are generally considered a transitional state between terrestrial-aquatic habitats and provide numerous vital ecosystem services to mankind. Wetlands are under a tremendous pressure due to growing human interference, urbanization, conventional agriculture, industrial expansions, and overexploitation of ecological services. The Keoladeo National Park (KNP) is a manmade wetland, world heritage site and a designated Ramsar site in India, widely known for its avian biodiversity. Due to insufficient amount of water supply and widespread invasion of Prosopis juliflora, notable spatio-temporal changes are observed in the land cover affecting habitat quality of the park. The present study is designed to highlight the importance of very high-resolution satellite data for characterization of the wetland ecosystem. It assesses the spatio-temporal dynamics of land use/land cover (LULC) and habitat quality, a model built in the InVEST (Integrated Valuation of Ecosystem Services and Tradeoffs) tool, is utilized to analyze the effect of land cover changes and increase in P. juliflora on habitat quality in the park. The study concludes that drastic changes in LULC and rampant spread of P. juliflora have deteriorated the quality of habitat for bird species. Furthermore, it highlights importance of geospatial tools in mapping, monitoring, and managing wetland ecosystems.

Identification of Conservation Priority Zones Using Spatially Explicit Valued Ecosystem Services: A Case from the Indian Sundarbans.

Demarcation of conservation priority zones (CPZs) using spatially explicit models is the new challenge in ecosystem services (ESs) research. This study identifies the CPZs of the Indian Sundarbans by integrating 2 different approaches, that is, ESs and ecosystem health (EH). Five successive steps were followed to conduct the analysis: First, the ESs were estimated using biophysical and economic methods and a hybrid method (that combines biophysical and economic methods); second, the vigor-organization-resilience (VOR) model was used for estimating EH; third, the risk characterization value (RCV) of ESs was measured using the function of EH and ESs; fourth, Pearson correlation test was performed to analyze the interaction between ESs and EH components; and fifth, the CPZs were defined by considering 7 relevant components: ecosystem vigor, ecosystem organization, ecosystem resilience, RCV, EH, ESs, and the correlation between EH and ESs. Among the major ecoregions of the Sundarbans, the highest ESs value in economic terms is provided by the mangrove ecosystem (US$19 144.9 million per year). The highest conservation priority score was projected for the Gosaba block, which is dominated by dense mangrove forests. The estimated CPZs were found to be highly consistent with the existing biodiversity zonations. The outcome of this study could be a reference for environmentalists, land administrators, researchers, and decision makers to design relevant policies to protect the high values of the Sundarbans ecosystem. Integr Environ Assess Manag 2020;16:773-787. © 2020 SETAC.

Examining the effects of forest fire on terrestrial carbon emission and ecosystem production in India using remote sensing approaches.

Remote sensing techniques are effectively used for measuring the overall loss of terrestrial ecosystem productivity and biodiversity due to forest fires. The current research focuses on assessing the impacts of forest fires on terrestrial ecosystem productivity in India during 2003-2017. Spatiotemporal changes of satellite remote sensing derived burn indices were estimated for both fire and normal years to analyze the association between forest fires and ecosystem productivity. Two Light Use Efficiency (LUE) models were used to quantify the terrestrial Net Primary Productivity (NPP) of the forest ecosystem using the open-source and freely available remotely sensed data. A novel approach (delta NPP/delta burn indices) is developed to quantify the effects of forest fires on terrestrial carbon emission and ecosystem production. During 2003-2017, the forest fire intensity was found to be very high (>2000) across the eastern Himalayan hilly region, which is mostly covered by dense forest and thereby highly susceptible to wildfires. Scattered patches of intense forest fires were also detected in the lower Himalayan and central Indian states. The spatial correlation between the burn indices and NPP were mainly negative (-0.01 to -0.89) for the fire-prone states as compared to the other neighbouring regions. Additionally, the linear approximation between the burn indices and NPP showed a positive relation (0.01 to 0.63), suggesting a moderate to high impact of the forest fires on the ecosystem production and terrestrial carbon emission. The present approach has the potential to quantify the loss of ecosystem productivity due to forest fires.

Modelling Agriculture, Forestry and Other Land Use (AFOLU) in response to climate change scenarios for the SAARC nations.

Agriculture and forestry are the two major land use classes providing sustenance to the human population. With the pace of development, these two land use classes continue to change over time. Land use change is a dynamic process under the influence of multiple drivers including climate change. Therefore, tracing the trajectory of the changes is challenging. The artificial neural network (ANN) has successfully been applied for tracing such a dynamic process to capture nonlinear responses. We test the application of the multilayer perceptron neural network (MLP-NN) to project the future Agriculture, Forestry and Other Land Use (AFOLU) for the year 2050 for the South Asian Association for Regional Cooperation (SAARC) nations which is a geopolitical union of Afghanistan, Bangladesh, Bhutan, India, Nepal, Maldives, Pakistan and Sri Lanka. The Intergovernmental Panel on Climate Change (IPCC) and Food and Agriculture Organization (FAO) use much frequently the term 'AFOLU' in their policy documents. Hence, we restricted our land use classification scheme as AFOLU for assessing the influence of climate change scenarios of the IPCC fifth assessment report (RCP 2.6, RCP 4.5, RCP 6.0 and RCP 8.5). Agricultural land would increase in all the SAARC nations, with the highest increase in Pakistan and Maldives; moderate increase in Afghanistan, India and Nepal; and the least increase in Bangladesh, Bhutan and Sri Lanka. The forestry land use will witness a decreasing trend under all scenarios in all of the SAARC nations with varying levels of changes. The study is expected to assist planners and policymakers to develop nations' specific strategy to proportionate land use classes to meet various needs on a sustainable basis.

Responses of ecosystem services to natural and anthropogenic forcings: A spatial regression based assessment in the world's largest mangrove ecosystem.

Most of the Earth's Ecosystem Services (ESs) have experienced a decreasing trend in the last few decades, primarily due to increasing human dominance in the natural environment. Identification and categorization of factors that affect the provision of ESs from global to local scales are challenging. This study makes an effort to identify the key driving factors and examine their effects on different ESs in the Sundarbans region, India. We carry out the analysis following five successive steps: (1) quantifying biophysical and economic values of ESs using three valuation approaches; (2) identifying six major driving forces on ESs; (3) categorizing principal data components with dimensionality reduction; (4) constructing multivariate regression models with variance partitioning; (5) implementing six spatial regression models to examine the causal effects of natural and anthropogenic forcings on ESs. Results show that climatic factors, biophysical factors, and environmental stressors significantly affect the ESs. Among the six driving factors, climate factors are highly associated with the ESs variation and explain the maximum model variances (R2 = 0.75-0.81). Socioeconomic (R2 = 0.44-0.66) and development (R2 = 27-0.44) factors have weak to moderate effects on the ESs. Furthermore, the joint effects of the driving factors are much higher than their individual effects. Among the six spatial regression models, Geographical Weighted Regression (GWR) performs the most accurately and explains the maximum model variances. The proposed hybrid valuation method aggregates biophysical and economic estimates of ESs and addresses methodological biases existing in the valuation process. The presented framework can be generalized and applied to other ecosystems at different scales. The outcome of this study could be a reference for decision-makers, planners, land administrators in formulating a suitable action plan and adopting relevant management practices to improve the overall socio-ecological status of the region.

Land use and climate change impacts on distribution of plant species of conservation value in Eastern Ghats, India: a simulation study.

Effective monitoring of the current status of species distributions and predicting future distributions are very important for conservation practices at the ecosystem and species levels. The human population, land use, and climate are important factors that influence the distributions of species. Even though future simulations have many uncertainties, such studies can provide a means of obtaining species distributions, range shifts, and food production and help mitigation and adaptation planning. Here, we simulate the population, land use/land cover and species distributions in the Eastern Ghats, India. A MaxEnt species distribution model was used to simulate the potential habitats of a group of endemic (28 species found in this region) and rare, endangered, and threatened (RET) (22 species found in this region) plant species on the basis of IPCC AR5 scenarios developed for 2050 and 2070. Simulations of populations in 2050 indicate that they will increase at a rate of 1.12% relative to the base year, 2011. These increases in population create a demand for more land for settlement and food productions. Land use land cover (LULC) simulations show an increase in built-up land from 3665.00 km2 in 2015 to 3989.56 km2 by 2050. There is a minor increase of 0.04% in the area under agriculture in 2050 compared with 2015. On the other hand, the habitat simulations show that the combined effects of climate and land use change have a greater influence on the decline of potential distributions of species. Climate change and the prevailing rate of LULC change will reduce the extents of the habitats of endemic and RET species (~ 60% and ~ 40%, respectively). The Eastern Ghats have become extensively fragmented due to human activities and have become a hotspot of endemic and RET species loss. Climate and LULC change will enhance the species loss and ecosystem services.

Ecosystem service value assessment of a natural reserve region for strengthening protection and conservation.

Ecosystem Services (ESs) refer to the direct and indirect contributions of ecosystems to human well-being and subsistence. Ecosystem valuation is an approach to assign monetary values to an ecosystem and its key ecosystem goods and services, generally referred to as Ecosystem Service Value (ESV). We have measured spatiotemporal ESV of 17 key ESs of Sundarbans Biosphere Reserve (SBR) in India using temporal remote sensing (RS) data (for years 1973, 1988, 2003, 2013, and 2018). These mangrove ecosystems are crucial for providing valuable supporting, regulatory, provisioning, and cultural ecosystem services. We have adopted supervised machine learning algorithms for classifying the region into different ecosystem units. Among the used machine learning models, Support Vector Machine (SVM) and Random Forest (RF) algorithms performed the most accurate and produced the best classification estimates with maximum kappa and an overall accuracy value. The maximum ESV (derived from both adjusted and non-adjusted units, million US$ year-1) is produced by mangrove forest, followed by the coastal estuary, cropland, inland wetland, mixed vegetation, and finally urban land. Out of all the ESs, the waste treatment (WT) service is the dominant ecosystem service of SBR. Additionally, the mangrove ecosystem was found to be the most sensitive to land use and land cover changes. The synergy and trade-offs between the ESs are closely associated with the spatial extent. Therefore, accurate estimates of ES valuation and mapping can be a robust tool for assessing the effects of poor decision making and overexploitation of natural resources on ESs.

Capturing forest dependency in the central Himalayan region: Variations between Oak (Quercus spp.) and Pine (Pinus spp.) dominated forest landscapes.

Our study explores the nexus between forests and local communities through participatory assessments and household surveys in the central Himalayan region. Forest dependency was compared among villages surrounded by oak-dominated forests (n = 8) and pine-dominated forests (n = 9). Both quantitative and qualitative analyses indicate variations in the degree of dependency based on proximity to nearest forest type. Households near oak-dominated forests were more dependent on forests (83.8%) compared to households near pine-dominated forests (69.1%). Forest dependency is mainly subsistence-oriented for meeting basic household requirements. Livestock population, cultivated land per household, and non-usage of alternative fuels are the major explanatory drivers of forest dependency. Our findings can help decision and policy makers to establish nested governance mechanisms encouraging prioritized site-specific conservation options among forest-adjacent households. Additionally, income diversification with respect to alternate livelihood sources, institutional reforms, and infrastructure facilities can reduce forest dependency, thereby, allowing sustainable forest management.

Mapping forests in monsoon Asia with ALOS PALSAR 50-m mosaic images and MODIS imagery in 2010.

Extensive forest changes have occurred in monsoon Asia, substantially affecting climate, carbon cycle and biodiversity. Accurate forest cover maps at fine spatial resolutions are required to qualify and quantify these effects. In this study, an algorithm was developed to map forests in 2010, with the use of structure and biomass information from the Advanced Land Observation System (ALOS) Phased Array L-band Synthetic Aperture Radar (PALSAR) mosaic dataset and the phenological information from MODerate Resolution Imaging Spectroradiometer (MOD13Q1 and MOD09A1) products. Our forest map (PALSARMOD50 m F/NF) was assessed through randomly selected ground truth samples from high spatial resolution images and had an overall accuracy of 95%. Total area of forests in monsoon Asia in 2010 was estimated to be ~6.3 × 10(6 )km(2). The distribution of evergreen and deciduous forests agreed reasonably well with the median Normalized Difference Vegetation Index (NDVI) in winter. PALSARMOD50 m F/NF map showed good spatial and areal agreements with selected forest maps generated by the Japan Aerospace Exploration Agency (JAXA F/NF), European Space Agency (ESA F/NF), Boston University (MCD12Q1 F/NF), Food and Agricultural Organization (FAO FRA), and University of Maryland (Landsat forests), but relatively large differences and uncertainties in tropical forests and evergreen and deciduous forests.

Geospatial quantification and analysis of environmental changes in urbanizing city of Kolkata (India).

Over the past five decades, the fragile wetland ecosystem surrounding the city of Kolkata has witnessed extensive changes in the name of urban development. In this study, we elaborate relationships among biophysical parameters and land surface temperature (LST) in Kolkata city and nearby surrounding areas where rapid urbanization has occurred. LST and associated surface physical characteristics were assessed using Landsat images acquired for the years 1989, 2006, and 2010. The satellite data was used to study the spatiotemporal urban footprint and correlation among normalized difference vegetation index (NDVI), normalized difference built-up index (NDBI), normalized difference water index (NDWI) and LST. Land use land cover (LULC) maps prepared using supervised classification had overall accuracy of 90, 88, and 86 % and kappa coefficient of 0.8726, 0.8455, and 0.8212 for 1989, 2006, and 2010, respectively. The spatial expansion as a consequence of increasing urban population is 108.94 km(2) over past two decades. The urban built-up in and around the city extends up to 88.71 km(2) in 1989, 144.64 km(2) in 2006, and 197.65 km(2) in 2010. These changes have attributed in elevating surface temperature in the study region. Analysis of biophysical parameters shows LST and NDBI having a positive correlation, LST and NDVI having negative correlation, while NDBI and NDWI having a perfectly negative correlation. Satellite estimated temperatures of the surface show a warming trend evident from increasing mean surface temperature values from 27.36 °C in 1989 to 30.025 °C in 2006 and 33.023 °C in 2010. The magnitude and extent of the estimates of LST are consistent with the urbanization pattern throughout the city and adjoining areas.

Spatio-temporal footprints of urbanisation in Surat, the Diamond City of India (1990-2009).

Urbanisation is a ubiquitous phenomenon with greater prominence in developing nations. Urban expansion involves land conversions from vegetated moisture-rich to impervious moisture-deficient land surfaces. The urban land transformations alter biophysical parameters in a mode that promotes development of heat islands and degrades environmental health. This study elaborates relationships among various environmental variables using remote sensing dataset to study spatio-temporal footprint of urbanisation in Surat city. Landsat Thematic Mapper satellite data were used in conjugation with geo-spatial techniques to study urbanisation and correlation among various satellite-derived biophysical parameters, [Normalised Difference Vegetation Index, Normalised Difference Built-up Index, Normalised Difference Water Index, Normalised Difference Bareness Index, Modified NDWI and land surface temperature (LST)]. Land use land cover was prepared using hierarchical decision tree classification with an accuracy of 90.4 % (kappa = 0.88) for 1990 and 85 % (kappa = 0.81) for 2009. It was found that the city has expanded over 42.75 km(2) within a decade, and these changes resulted in elevated surface temperatures. For example, transformation from vegetation to built-up has resulted in 5.5 ± 2.6 °C increase in land surface temperature, vegetation to fallow 6.7 ± 3 °C, fallow to built-up is 3.5 ± 2.9 °C and built-up to dense built-up is 5.3 ± 2.8 °C. Directional profiling for LST was done to study spatial patterns of LST in and around Surat city. Emergence of two new LST peaks for 2009 was observed in N-S and NE-SW profiles.

Landscape approach for quantifying land use land cover change (1972-2006) and habitat diversity in a mining area in Central India (Bokaro, Jharkhand).

The rate and intensity of land use land cover (LULC) change has increased considerably during the past couple of decades. Mining brings significant alterations in LULC specifically due to its impact on forests. Parts of Central India are well endowed with both forests and minerals. Here, the conflict between human interests and nature has intensified over time. Monitoring and assessment of such conflicts are important for land management and policy making. Remote sensing and Geographical Information System have the potential to serve as accurate tools for environmental monitoring. Understanding the importance of landscape metrics in land use planning is challenging but important. These metrics calculated at landscape, class, and patch level provide an insight into changing spatiotemporal distribution of LULC and ecological connectedness. In the present study, geospatial tools in conjunction with landscape metrics have been used to assess the impact of coal mining on habitat diversity. LULC maps, change detection analysis, and landscape metrics have been computed for the four time periods (1972, 1992, 2001, and 2006). There has been a significant decline in forest cover especially of the Sal-mixed forests, both in area as well as quality, due to flouted mining regulations. Reclamation of mined lands has also been observed in some of the areas since 2001.