Intensive anthropogenic activities in high-groundwater coal basins (HGCBs) have dramatically altered land use/land cover (LULC) patterns, exerting profound impacts on regional ecosystem carbon storage. In this study, conducted in the Yanzhou Coalfield, China, we developed a refined water body classification scheme that divides water bodies into natural, artificial, seasonal, and perennial types to better capture the ecological heterogeneity of water bodies within mining-induced subsidence areas, which is often overlooked in conventional assessments. By coupling the PLUS and InVEST models, we quantified carbon storage dynamics driven by LULC changes from 2005 to 2020 and projected carbon storage levels for 2035 under four scenarios: historical trend (HT), food security (FS), ecological restoration (ER), and coordinated development (CD). The primary findings of this study are as follows: (1) Significant differences in carbon density existed among water body types, in the order natural water bodies > seasonal water bodies > perennial water bodies > artificial water bodies. Neglecting these differences can introduce systematic biases in carbon storage assessment. (2) Cropland area decreased substantially from 2005 to 2020, while built-up land and water bodies expanded continuously. (3) Intensive LULC transitions resulted in a 12.77% decline in total carbon storage (equivalent to 2.38 × 105Mg), primarily driven by the conversion of high-carbon-density cropland to built-up land and subsidence water bodies due to coal mining activities. (4) Projections for 2035 indicate that total carbon storage would decline by an additional 11.67% under the HT scenario. In contrast, the FS, ER, and CD scenarios effectively mitigated losses, with declines of 6.15%, 5.05%, and 3.13%, respectively. The CD scenario achieved the best performance through synergistic optimization of carbon sequestration, food security, and ecological restoration objectives. These findings provide critical insights for integrating carbon sequestration objectives into land management practices in coal mining areas, thereby supporting land use optimization and climate change mitigation efforts.

Against the backdrop of increasing global climate change and human activities, the impact of land use change on carbon storage has become a significant issue in ecosystem service research, particularly in the ecologically fragile arid regions, which poses serious challenges to climate regulation and ecological sustainability. To address this issue, this study focuses on the arid region of northwest China, systematically analyzing the spatiotemporal changes in land use patterns from 2000 to 2020 and their impact on regional carbon storage. The aim is to explore the potential impacts of land use changes on the evolution of carbon storage functions under three scenarios: natural development, ecological protection, and economic development. This research employs the InVEST model to assess the spatiotemporal distribution characteristics of carbon storage, utilizes the PLUS model to perform multi-scenario simulations, and incorporates the Geodetector method to identify the main driving factors behind the spatial differentiation of carbon storage. The results indicate that: ① Over the past 20 years, land use changes in the northwest arid region have been significant, with continuous expansion of cultivated and construction lands, while grassland areas have first decreased and then increased, and forest and unused lands have substantially degraded. Notably, the increase in cultivated land area is particularly prominent, whereas the areas of forest and unused lands have shown a declining trend； grassland experienced a continuous reduction from 2000 to 2015, followed by a recovery from 2015 to 2020. ② In terms of carbon storage trends, carbon storage in the northwest arid region increased annually from 2000 to 2020, reaching 9.01×109 t in 2020, an increase of over 1.68×108 t compared to that in 2000, with grassland consistently contributing the most to carbon storage. ③ In the multi-scenario simulations, the carbon storage levels for 2030 under the natural development, ecological protection, and economic development scenarios were projected to be 9.10×109 t, 9.11×109 t, and 9.09×109 t, respectively, highlighting significant impacts of different development paths on carbon storage. Among these, the ecological protection scenario showed the greatest increase in carbon storage, indicating that reinforcing ecological protection measures can effectively enhance the carbon storage capacity of the region. ④ NDVI has been identified as a key driving factor explaining the spatial differentiation of carbon storage in the northwest arid region, with a q value of 0.432 7. The interaction of NDVI with factors such as soil sand content and climate showed a significant two-factor enhancement effect, improving the explanation of spatial differences in carbon storage. This research reveals the close coupling relationship between land use changes and carbon storage functions, providing scientific evidence for ecological protection, optimal allocation of land resources, and strategies for enhancing carbon sinks in arid regions.

Universities worldwide are increasingly committing to carbon neutrality; however, most institutional climate strategies treat operational emissions forecasting and ecosystem-based carbon sequestration as separate analytical domains, leading to inconsistencies in accounting boundaries, temporal alignment, and verification practices. This study develops and demonstrates an integrated LEAP–InVEST framework that explicitly links energy-system modeling with spatial ecosystem carbon accounting within a unified monitoring, reporting, and verification (MRV)-aligned structure. The framework combines the Low Emissions Analysis Platform (LEAP) for scenario-based greenhouse gas emissions modeling with the Integrated Valuation of Ecosystem Services and Tradeoffs (InVEST) model for spatial carbon storage assessment. A key methodological contribution lies in reconciling emission flows and carbon stock changes by converting carbon stock variations into annualized removal flows, thereby enabling consistent estimation of gross emissions, carbon removals, and net emissions while avoiding double counting across scopes. Using a university campus in Taiwan as a case study, a baseline inventory was established following ISO 14064-1 standards, and future emissions trajectories were simulated under Business-as-Usual and mitigation pathways through 2040. In parallel, land-use and land-cover data were used to quantify historical and projected carbon stocks across forest, grassland, agricultural, and built-up areas. Results indicate that electricity consumption constitutes the dominant emissions source, and that energy efficiency improvements, photovoltaic deployment, and green power procurement provide the largest mitigation potential. Although ecosystem carbon stocks remain substantial, their annual sequestration capacity offsets only a limited portion of projected emissions, reinforcing the importance of prioritizing emissions reduction before applying nature-based removals. The proposed framework provides a transferable methodological approach for institutional carbon neutrality planning by integrating emissions reduction and carbon sequestration within a coherent analytical system. By aligning energy modeling, ecosystem dynamics, and MRV principles, the framework enhances the transparency, credibility, and robustness of net-zero pathway assessment and is applicable to universities and compact urban systems seeking data-driven and verifiable decarbonization strategies.

Economic and environmental impact assessment of carbon storage in Malaysia

Regional carbon emissions are closely related to land use composition and its pattern. Optimizing the spatial distribution of land use is pivotal for reducing carbon emissions and enhancing carbon sequestration, thereby contributing to the achievement of the "carbon neutrality" goal in 2060. This study established a future land use simulation （FLUS） model of Guangzhou based on the land use change in 2015-2020 and optimally allocated the land uses in Guangzhou using the carbon emission coefficient method and the linear programming model, which was triggered by the goal of carbon neutrality. The integrated valuation of ecosystem services and trade-offs （InVEST） model was employed to assess the carbon emission and carbon storage of land uses in 2015-2020 and the optimized land use pattern for 2060 in Guangzhou. Adaptive strategies and suggestions were proposed to achieve carbon neutrality in Guangzhou. The findings are as follows： ① During the period of 2015-2020, land use changes in Guangzhou were characterized by the outward and infill expansion of urban land, occupying the surrounding farmland and ecological land. This resulted in a reduction of 2.4×106 t carbon emissions and 1.26×105 t carbon stock, with 89.02% of the land achieving carbon balance. ② Land use optimization under the 2060 carbon neutrality goal could greatly restrict land use conversion. The built-up land could increase by 17 038 hm2, predominantly through extension, while ecological land would increase by 6 521.68 hm2, mainly through the integration of small patches. Farmland would decrease by 23 446 hm2. ③ With the carbon neutrality target, the 2060 land use could reduce net carbon emission, accounting for 5.5% of that in 2015, and bolster carbon stock by an additional 1.12×105 t. The spatial effects of carbon emissions could be weakened, and 96.89% of the land would achieve carbon balance. This study contributes a robust scientific foundation, using Guangzhou as a reference towards a low-carbon urbanization, thereby promoting the attainment of carbon neutrality.

ABSTRACT The accurate estimation of forest canopy height (FCH) is significant for research on carbon storage and climate change. Extrapolating FCH samples at the plot scale to the global or regional level using spatially contiguous remote sensing images is currently a common method for large-scale vegetation attribute mapping. However, the extrapolation process results in a lower FCH estimation accuracy due to signal saturation, a lack of horizontal stand distribution, and incomplete seasonal information expression. Here, a new FCH remote sensing estimation model was created using the seasonal rhythms of dominant tree species (pine, oak, walnut, and other species) in the eastern part of Dali Bai Autonomous Prefecture, Yunnan Province, China. First, a spectral curve model was used to extract seasonal information from the integrated Landsat and Sentinel-2 imagery. Following this, the seasonal rhythm and support vector machine (SVM) were adopted to identify the forest dominant tree species. Integrating active and passive remote sensing features and seasonal rhythm variables, multiple random forest-based FCH estimation models of different forest dominant tree species were constructed to reveal the distribution of FCH. Our results determined a reliable overall classification accuracy of dominant tree species at 0.92, indicating that the classification could be used for subsequent FCH modeling. The validation R 2 of the FCH model using active and passive remote sensing variables with the random forest algorithm was only 0.37, while the R 2 value increased to 0.57 when the seasonal rhythm information was included. After the FCH model was established for each forest dominant tree species, the validation R 2 further increased to 0.70. These results revealed that the seasonal rhythm information and forest tree species composition contribute to reliable FCH distribution mapping. This study provides a robust framework for improving FCH estimation, offering valuable insights for forest management strategies and enhancing carbon storage assessment.

Optimizing the spatial pattern of its carbon storage is of great significance for increasing the carbon storage capacity of regional ecosystem and maintaining regional carbon balance. Although the existing research has achieved remarkable results in regional carbon storage assessment and multi-scenario simulation studies, there are still obvious deficiencies in determining specific carbon storage optimization areas for developed regions and formulating targeted low-carbon development strategies. Taking the economically developed Jiangsu section of the Yangtze River Basin (JS-YRB) as an example, combined with InVEST and PLUS models, the carbon storage and its spatial distribution pattern of the study area in 2030 were predicted under three different scenarios: natural development, cropland protection and ecological protection. The pattern of carbon storage in the study area was optimized by a Bayesian belief network (BBN) with decision optimization ability. The results showed that: (1) From 2000 to 2020, the carbon storage in the study area exhibited a decreasing trend, with a total reduction of 47.98 × 106 t. The primary reason for these decreases was the conversion of cropland and forest land to built-up land. (2) In 2030, under the ecological protection scenario, the carbon storage in the study area would be 390.58 × 106 t, showing an upward trend, while under the other two scenarios, the carbon storage would show a downward trend. (3) Key variables and key state subsets were selected by BBN, and the study area would be divided into four types of optimal zones: ecological protection area, cropland protection area, water conservation area and economic construction area. The findings can provide a reference for the sustainable development of land use within the watershed and contribute to advancing the watershed's efforts toward achieving the carbon neutrality goals.

Mangrove ecosystems are globally recognized as critical blue carbon reservoirs due to their exceptional capacity to sequester and store carbon in both biomass and soils. Consequently, the objective of this paper is to investigate the Spatial Variation in Carbon Storage in Mangrove Ecosystems of the Niger Delta: Insights from the Bodo Mangrove Forest in Rivers State, Nigeria using appropriate standard methods. Results reveal significant spatial heterogeneity, with total ecosystem carbon stocks ranging from 414 to 972 Mg C ha⁻¹, stored in soils (78–94%). Interior basin zones exhibited the highest carbon stocks, attributed to mature vegetation, favourable hydrological regimes, and sediment characteristics, whereas disturbed areas showed substantial carbon depletion linked to historical oil spills. Multivariate analyses identified tidal inundation frequency, soil texture, and species dominance-particularly Rhizophora racemosa-as key drivers of carbon distribution. These findings underscore the critical importance of conserving intact mangrove habitats and inform restoration strategies aimed at maximizing carbon sequestration. The study contributes valuable spatially explicit data to enhance carbon accounting accuracy and supports the integration of Niger Delta mangroves into global climate mitigation frameworks. Carbon storage assessment in mangrove ecosystems represents a critical frontier in climate change mitigation research, with Nigerian mangroves emerging as significant yet understudied blue carbon repositories. This study presents a comprehensive analysis of spatial heterogeneity in carbon sequestration across the Bodo mangrove forest in the Niger Delta region, revealing substantial variations in carbon stocks correlated with ecological zonation, anthropogenic disturbance patterns, and hydrogeomorphic characteristics.

Land-use change exerts a profound influence on ecosystem services (ES), and accurately assessing its spatiotemporal dynamics is essential for achieving regional sustainability. Taking Shandong Province as a case study, this research integrates the PLUS and InVEST models to simulate the impacts of land-use changes on carbon storage and habitat quality in Shandong Province between 2000 and 2020, and to project their dynamics under different scenarios for 2030. The InVEST (Integrated Valuation of Ecosystem Services and Tradeoffs) model was employed to reassess variations in carbon storage (CS) and habitat quality (HQ). The main findings are as follows: (1) Cultivated land decreased by 12.3%, while construction land expanded by 51.04%, predominantly replacing farmland and forested areas, resulting in a distinct spatial pattern characterized by an “urbanized east and agricultural west.” (2) Carbon storage declined by approximately 63 million tons, primarily due to urban expansion. (3) Habitat quality experienced a 3.6% decrease, with significant ecological fragmentation identified in the central mountainous regions and the Yellow River Delta, driven by intensified urbanization and agricultural activities. (4) Future scenario simulations indicate that under the ecological conservation scenario, carbon storage could increase by 12.5% and habitat quality could reach 0.572 by 2040; in contrast, the natural development scenario suggests ongoing degradation. These findings highlight the trade-offs between land development and ecosystem services, emphasizing the necessity of reinforcing ecological zoning, compensation mechanisms, and the establishment of ecological corridors. This study provides a scientific basis for advancing sustainable land-use planning and ecosystem management.Supplementary InformationThe online version contains supplementary material available at 10.1038/s41598-025-25097-y.

In the pursuit of sustainable urban planning, integrating land use simulation with carbon storage assessment is crucial for achieving the “dual carbon” goals. This study focuses on the Fuzhou Metropolitan Area, utilizing land use data from 2000, 2010, and 2020. By establishing three future development scenarios—natural, urban, and dual-carbon target scenarios—based on the “Fuzhou Metropolitan Area Development Plan,” this research employ the Patch-generating Land Use Simulation (PLUS) and the Integrated Valuation of Ecosystem Services and Trade-offs (InVEST) models. The analysis reveals that from 2000 to 2020, the areas of cultivated land, forest land, grassland, and water bodies decreased, while construction land and bare land increased. Notably, the nighttime lighting factor significantly impacts land use changes, with elevation playing a crucial role in changes to water bodies and bare land. Under natural and urban development scenarios, carbon storage exhibits a downward trend, whereas the dual-carbon target scenario limits construction land expansion and reverses this trend, resulting in increased carbon storage. Based on these insights, this study proposes a three-stage urban planning strategy: strengthening carbon assessment in the early stages, fostering cross-departmental collaboration during implementation, and ensuring dynamic monitoring and adaptive adjustments in the later stages. This approach aims to harmonize urban development with ecological conservation, thereby maximizing economic and ecological benefits and supporting the achievement of the “dual carbon” policy goals.Supplementary InformationThe online version contains supplementary material available at 10.1038/s41598-025-13961-w.

Land use/cover changes lead to a decrease in carbon storage in arid regions-a case study of Northwest China.

Optimization of remote sensing estimation model for biomass of rubber plantations from the perspective of multi-source feature fusion

Carbon storage assessment using machine learning approaches in Qadirpur Gas Field, Pakistan

Multi-scenario simulation and carbon storage assessment of land use in a multi-mountainous city

The land cover in the coastal zone is characterized by frequent changes, fragmented landscape and strong spatial heterogeneity, which makes accurate assessment and analysis of coastal ecosystem carbon storage challenging. This study developed a coastal ecosystem carbon storage assessment framework by integrating Landsat time-series analysis with the Integrated Valuation of Ecosystem Services and Tradeoffs (InVEST) model. First, using the GEE cloud platform and Landsat long-term satellite remote sensing data, this study applied median compositing algorithms to mitigate the impact of periodic tidal inundation on land boundaries. Second, by integrating multiple feature parameters and utilizing the random forest method, accurate information on land use/cover change was obtained. Subsequently, carbon density parameters were determined, and coastal ecosystem carbon storage was assessed using the InVEST model. Finally, a spatiotemporal pattern analysis of coastal ecosystem carbon storage in Hangzhou Bay over nearly four decades was conducted. The findings yielded the subsequent outcomes: (1) The random forest algorithm integrated multiple feature parameters is stable, and can extract LUCC information accurately. (2) The overall coastal ecosystem carbon storage of Hangzhou Bay, China, witnessed a decline over the preceding four decades, dropping from 108.15 Mt in 1985 to 82.47 Mt in 2023. (3) The decrease of vegetation area and the expansion of build-up area are the main reasons for the change of carbon storage. This study furnishes valuable data support to underpin the strategic governance of land resources in the Hangzhou Bay region, while the resultant carbon storage dataset holds critical ramifications for regional sustainable development.

Surface mining activities cause severe disruption to ecosystems, resulting in the substantial destruction of surface vegetation, the loss of soil organic carbon stocks, and a decrease in the ecosystem’s ability to sequester carbon. The ecological restoration of mining areas has been found to significantly enhance the carbon storage capacity of ecosystems. This study evaluated ecological restoration strategies in Chongqing’s Tongluo Mountain mining area by integrating GF-6 satellite multispectral data (2 m panchromatic/8 m multispectral resolution) with ground surveys across 45 quadrats to develop a quadratic regression model based on vegetation indices and the field-measured biomass. The methodology quantified carbon storage variations among engineered restoration (ER), natural recovery (NR), and unmanaged sites (CWR) while identifying optimal vegetation configurations for karst ecosystems. The methodology combined the high-spatial-resolution satellite imagery for large-scale vegetation mapping with field-measured biomass calibration to enhance the quantitative accuracy, enabling an efficient carbon storage assessment across heterogeneous landscapes. This hybrid approach overcame the limitations of traditional plot-based methods by providing spatially explicit, cost-effective monitoring solutions for mining ecosystems. The results demonstrate that engineered restoration significantly enhances carbon sequestration, with the aboveground vegetation biomass reaching 5.07 ± 1.05 tC/ha, a value 21% higher than in natural recovery areas (4.18 ± 0.23 tC/ha) and 189% greater than at unmanaged sites (1.75 ± 1.03 tC/ha). In areas subjected to engineered restoration, both the vegetation and soil carbon storage showed an upward trend, with soil carbon sequestration being the primary form, contributing to 81% of the total carbon storage, and with engineered restoration areas exceeding natural recovery and unmanaged zones by 17.6% and 106%, respectively, in terms of their soil carbon density (40.41 ± 9.99 tC/ha). Significant variations in the carbon sequestration capacity were observed across vegetation types. Bamboo forests exhibited the highest carbon density (25.8 tC/ha), followed by tree forests (2.54 ± 0.53 tC/ha), while grasslands showed the lowest values (0.88 ± 0.52 tC/ha). For future restoration initiatives, it is advisable to select suitable vegetation types based on the local dominant species for a comprehensive approach.

Land use and land cover change (LULCC) is a key driver of carbon storage changes, especially in complex coastal ecosystems such as the Yellow River Delta (YRD), which is jointly influenced by climate change and resource development. The compounded effects of sea-level rise (SLR) and land subsidence (LS) are particularly prominent. This study is the first to integrate the dual impacts of SLR and LS into a unified framework, using three climate scenarios (SSP1–26, SSP2–45, SSP5–85) provided in the Intergovernmental Panel on Climate Change (IPCC) Sixth Assessment Report (AR6), along with LS monitoring data, to comprehensively assess future inundation risks. Building on this, and taking into account land use and ecological protection policies in the YRD, three strategic scenarios—Ecological Protection Scenario (EPS), Natural Development Scenario (NDS), and Economic Growth Scenario (EGS)—are established. The PLUS and InVEST models are used to jointly simulate LULCC and carbon storage changes across these scenarios. Unlike previous studies focusing on single driving factors, this research innovatively develops a dynamic simulation system for LULCC and carbon storage driven by the SLR-LS compound effects, providing scientific guidance for land space development and coastal zone planning in vulnerable coastal areas, while enhancing carbon sink potential. The results of the study show the following: (1) Over the past 30 years, the land use pattern of the YRD has generally extended toward the sea, with land use transitions mainly from grasslands (the largest reduction: 1096.20 km2), wetlands, reservoirs and ponds, and paddy fields to drylands, culture areas, construction lands, salt pans, and tidal flats. (2) Carbon storage in the YRD exhibits significant spatial heterogeneity. Low-carbon storage areas are primarily concentrated in the coastal regions, while high-carbon storage areas are mainly found in grasslands, paddy fields, and woodlands. LULCC, especially the conversion of high carbon storage ecosystems to low carbon storage uses, has resulted in an overall net regional carbon loss of 2.22 × 106 t since 1990. (3) The risk of seawater inundation in the YRD is closely related to LS, particularly under low sea-level scenarios, with LS playing a dominant role in exacerbating this risk. Under the EGS, the region is projected to face severe seawater inundation and carbon storage losses by 2030 and 2060.

Forests play an important role in the global carbon cycle, making accurate assessments of carbon dynamics essential for effective forest management and climate change mitigation strategies. This research examines the spatiotemporal patterns of carbon storage and sequestration (CSS) in forests' aboveground biomass using satellite data, machine learning (Support Vector Machines), carbon modelling and spatial statistics. The methodology follows a two-step classification process: (i) binary forest classification and (ii) forest type classification, mapping seven forest types within two main categories - Broadleaves ( Quercus suber, Quercus ilex, Eucalyptus sp., and other species) and Coniferous ( Pinus pinaster, Pinus pinea, and other species). We analyzed the relationship between forest type and CSS at the Nomenclature of Territorial Units for Statistics (NUTS) III level and identified spatial clusters, outliers, and hot and cold spots of carbon sequestration at the municipal level across mainland Portugal. The broadleaved category demonstrated the highest classification accuracy in both years, decreasing slightly from 90.3 % in 2018 to 89 % in 2022, while the Coniferous group had the lowest accuracy, declining from 84.1 % in 2018 to 83.6 % in 2022. Anselin's Local Moran's I identified clusters of carbon sequestration, while the Getis-Ord Gi analysis confirmed these findings, revealing statistically significant hotspots of carbon sequestration in the northern and central regions and cold spots in the southern region. By providing insights at the sub-regional and municipal levels, this study offers a robust framework to support sustainable forest management and climate change mitigation strategies. Moreover, it can assist decision-makers in prioritizing natural capital, and developing nature-based solutions to tackle climate change and biodiversity loss.

Sacred groves are traditionally protected forest patches conserved by local communities through religious and cultural practices. While their ecological and cultural importance is well recognised, their role in climate change mitigation through carbon storage remains underexplored. This study assesses the carbon sequestration potential of selected sacred groves in and around the Union Territory of Puducherry, India. Three representative groves— Suriyanpet, Urani, and Kizhputhupet—were selected based on size, vegetation density, and disturbance levels. A non-destructive sampling approach using 20 m × 20 m quadrats was employed to estimate above- and belowground biomass using standard allometric equations, followed by carbon stock estimation through established biomass-to-carbon conversion factors. The results indicate that sacred groves function as significant carbon sinks despite fragmentation and increasing anthropogenic pressures. The study underscores the need to conserve and restore sacred groves as effective nature-based solutions for climate change mitigation, while also supporting biodiversity conservation and cultural heritage preservation.

Populus euphratica is an endangered species in desert riparian forests along the Tarim River in the arid region of northwestern China. Accurately estimating aboveground biomass is essential for assessing the growth status of P. euphratica , but is challenging as harvesting P. euphratica is strictly prohibited. In this study, we developed new allometric equations to estimate the woody aboveground biomass and volume of P. euphratica based on data collected along 200 km of the lower Tarim River. Our approach is nondestructive as it combines terrestrial laser scanning and quantitative structure modeling. We found that quantitative structure model effectively acquired tree 3D structures from high-density point clouds. The total volumes of the sample trees were found to range from 0.01 to 4.25 m 3 , and aboveground biomass from 3.07 to 1,997.41 kg, with 86% of the trees having a biomass below 500 kg. The biomass of large trees has been underestimated by previous allometric equations due to the limited sample sizes used to build the models. Our new allometric equations utilizing diameter at breast height and tree height ( AGB = 20.86 + 152.58 · H · DBH 2 ; Volume = 0.096 + 0.316 · H · DBH 2 ) showed less bias ( R 2 ≥ 0.94) since our data encompass not only small-sized but also large trees, resulting in a reduction of RMSE by 40% to 50% compared to previous models. When tree height was not available, diameter alone also provided high accuracy for estimating biomass ( AGB = 1,454.02 · DBH 2.054 ) and volume ( Volume = 0.15 · e 3.02 · DBH ), with R 2 values of 0.93. Our research provides a nondestructive method to accurately estimate the biomass of P. euphratica , contributing to improved forest conservation and carbon storage assessment in the desert riparian forest.