Restoration Strategies Research Articles

Effects of Restoration Strategies on the Ion Distribution and Transport Characteristics of Medicago sativa in Saline–Alkali Soil

Studying the distribution and transport dynamics of cations in plants is crucial for understanding their response mechanisms to saline–alkali stress conditions. However, our current understanding of how restoration measures affect cation distribution and transport in plants is surprisingly limited. To address this gap, we conducted a split-plot experiment using Medicago sativa L. cv. “Zhongmu No. 1” to investigate the combined effects of biological and chemical restoration measures—with bio-fertilizer as the primary zone and flue gas desulfurization (FGD) gypsum and with humic acid as the secondary zone—on soil properties, plant growth, and the content, distribution, and transport of cations in plants. The results revealed that bio-fertilizers exhibited positive effects on the plant growth, yield, and translocation of key ionic components to leaves. On the contrary, FGD gypsum with humic acid reduced the soil’s pH level, exchangeable sodium percentage (ESP), and sodium adsorption ratio (SAR) while increasing the contents of K+, Ca2+, and Mg2+ in the soil. The combination of bio-fertilizer, FGD gypsum, and humic acid increased the biomass and enhanced the translocation of Mg2+ to leaves. The distribution and transport of Mg2+ within the plant constituted pivotal elements for enhancing plant growth through restoration strategies. The application of bio-fertilizer, FGD gypsum, and humic acid reduced Na+ transport in M. sativa by enhancing the selective absorption of beneficial ions in leaves and by facilitating the transport of Ca2+ and Mg2+ from stems to the leaves. This, in turn, increases the salt tolerance of plants and promotes their growth. Our results offer new insights into the interactions among measures, soil, and plants in saline–alkali land restoration, providing practical solutions for the restoration of saline–alkali soil.

Adaptive shifts in plant traits associated with nitrogen removal driven by phytoremediation strategies in subtropical river restoration

Phytoremediation, which is commonly carried out through hydroponics and substrate-based strategies, is essential for the effectiveness of nature-based engineered solutions aimed at addressing excess nitrogen in aquatic ecosystems. However, the performance and mechanisms of plants involving nitrogen removal between different strategies need to be deeply understood. Here, this study employed in-situ cultivation coupled with static nitrogen tracing experiments to elucidate the influence of both strategies on plant traits associated with nitrogen removal. The results indicated that removal efficiencies in plants with substrate-based strategies for ammonium nitrogen and nitrate nitrogen were 30.51–71.11 % and 16.82–99.95 %, respectively, which were significantly higher than those with hydroponics strategies (25.98–58.18 % and 7.29–79.19 %, respectively). Similarly, the plant nitrogen uptake rates in the substrate-based strategy also generally showed higher levels compared to hydroponics strategies (P < 0.05). Meanwhile, the microorganisms-mediated nitrous oxide emission rates in the substrate-based strategy during summer (unamended: 0.00–0.58 μg/g/d; potential: 3.35–7.65 μg/g/d) were obviously lower than those in the hydroponics strategy (unamended: 2.23–11.70 μg/g/d; potential: 9.72–43.09 μg/g/d) (P < 0.05). Notably, analysis of similarity tests indicated that the influences of strategy on the above parameters generally surpass the effects attributable to interspecies plant differences, particularly during summer (R > 0, P < 0.05). Based on statistical and metagenomic analyses, this study revealed that these differences were driven by the stabilizing influence of substrate-based strategy on plant roots and enhancing synergistic interplay among biochemical factors within plant-root systems. Even so, phytoremediation strategies did not significantly alter the characteristics of plants with regards to their tendency towards ammonium nitrogen uptake (up to 87.68 %) and dissimilatory nitrate reduction to ammonium as primary biological pathway for nitrogen transformation which accounted for 53.66–96.47 % nitrate removal. In summary, this study suggested that the substrate-based strategy should be a more effective strategy for enhancing the nitrogen removal ability of plants in subtropical river restoration practices.

Evaluating the soil hydraulic functions of vegetation restoration in a subtropical hilly catchment: Insights from continuous soil moisture monitoring

The global trend of vegetation “greening” in the context of ecological restoration necessitates an urgent assessment of ecosystem services. As essential components of ecosystem services, the hydraulic functions of soil in infiltrating and retaining water following vegetation restoration remain unclear, especially in subtropical mountainous and hilly areas with complex topographies. From 2018 to 2021, soil moisture data collected at five-minute intervals were monitored for three restoration strategies in a hilly catchment of China’s Three Gorges Reservoir area. The restoration strategies included planted forest (PF) and natural restoration (naturally regenerated forest, NF; deforested pasture, DP). The soil moisture response to rainfall under these strategies was evaluated using several metrics, including the time difference between peak rainfall intensity and peak soil moisture response (Tp2p), cumulative infiltration, and occurrence frequency of preferential flow (PRF). The results showed that the average soil moisture content (SMC) of PF was significantly (p < 0.05) lower than that of NF and DP in both dry and wet seasons, regardless of upslope or downslope location. However, the topographic position affected the difference in average SMC between DP and NF. At the downslope location, the average SMC of DP (0.39 cm3 cm−3) was significantly higher than that of NF (0.33 cm3 cm−3). Conversely, at the upslope location, the average SMC of DP (0.27 cm3 cm−3) was lower than that of NF (0.30 cm3 cm−3). These findings suggested that PF had a lower amount of soil water storage than NF and DP, which was supported by the lowest cumulative infiltration in PF during storm events. The response of soil moisture to storms in PF (Tp2p = 3.1 h) was slower than that in NF (Tp2p = 1.9 h) and DP (Tp2p = 2.5 h). This was consistent with the lower occurrence frequency of preferential flow in PF (PRF = 19.2 %) than NF (PRF = 39.2 %) and DP (PRF = 32.9 %). Therefore, longer response time and less preferential flow indicated that the PF had a relatively poor soil moisture responsiveness to storm events. Accordingly, this study highlights the insufficiency of afforestation in soil water infiltration and retention compared to natural restoration, meriting consideration when assessing soil hydraulic functions in vegetation restoration areas.

Consideration on Legal Situation of Mine Ecological Restoration in China

Mine ecological restoration system is the guarantee of restoration practice and management. In the past half century, China's mining geological environment protection and restoration management system and land reclamation system have been established and continuously improved, but the current relevant legal system still has some defects, such as: no special legislation, implementations of ecological restoration settlement mechanism, two-in-one scheme compilation and reporting system, mining geological environment restoration t fund system are not smooth, and the binding force of legal responsibility is insufficient. In view of these problems, the revision of Mineral Resources Law, Regulations on Land Reclamation and Provisions on the Protection of the Geological Environment in Mines has become a practical necessity. In the revision, mine ecological restoration system can be constructed from the aspects of preparation and implementation supervision of mine ecological protection and restoration scheme, management of mine ecological restoration fund, ecological background investigation and monitoring, lifelong responsibility investigation, joint punishment and so on.

Participation of grassland species in various successional series in a temperate European region and implications for habitat management

Restoration strategies of semi-natural grasslands at various disturbed sites require an understanding of the resilience of these habitats and knowledge of their specific successional dynamics. Here, we assess (i) to what extent and under what circumstances grassland species can spontaneously establish and survive in various successional series, and (ii) whether some successional stages may provide surrogate habitats for typical grassland species. Besides grassland species, we evaluated the participation of other species groups, namely woodland, synanthropic and open wetland species, to demonstrate general successional trends. We studied 12 different successional series ca 60 years long in the Czech Republic, central Europe, and evaluated changes in species numbers and cover of species groups. Our results showed grassland species participated not only temporarily but also continuously during the first six decades of succession, as we had expected. Their decline in the late successional stages was only slightly visible despite the formation of woodland in most successional series. In some cases, the woodland was not completely closed and still allowed some grassland species to survive in the understorey. However, we found clear negative correlations between the participation of grassland and woodland species in the late successional stages. Woodland species generally increased, while synanthropic species clearly decreased during succession. Open wetland species slightly decreased, probably due to increasing site terrestrialisation in several successional series. The number and cover of dry grassland species decreased with increasing altitude, while those of mesic grassland species increased, indicating a decisive role of climate for their participation in succession. Successional trends discussed here represent general patterns but cannot be automatically applied in restoration without considering the specificity of each successional series. The establishment of grasslands in most successional series can be prevailingly maintained by management to reduce the establishment and expansion of woody species, otherwise grasslands only rarely persist in habitats under extreme site conditions or certain natural disturbances.

The relative impacts of vegetation, topography and weather on landscape patterns of burn severity in subtropical forests of southern China

Understanding the landscape patterns of burn severity is vital for managing fire-prone ecosystems. Relatively limited research has been done about fire and burn severity patterns in subtropical forests. Here, we derived the pre-fire forest type data from a global land-cover product at 30 m resolution based on time-series Landsat imageries. Using Landsat 8 OLI remote sensing imagery and field-based composite burn index (CBI), this study spatially mapped the burn severity of 27 forest fires in the subtropical forest ecosystems in southern China from 2017 to 2021. The landscape pattern of patches with different burn severity was quantified using landscape indices. In addition, factors influencing the patterns of burn severity across the landscape were determined using the Geodetector model. Burn severity of patches varied significantly over space. High burn severity was common in forest patches with low fragmentation, low patch density, and regular shape. In contrast, moderate and low burn severity was prevalent in patches with smaller patch size, high patch density, and complex shapes. Extensively burned forest patches were located at higher elevations, while more fragmented patches were located in gently sloping areas. Topographic factors were the most significant factors influencing variances in burn severity across the forest patches, followed by weather conditions. Compared to low elevation areas, vegetation types at the high elevation areas (dominated by Masson pine) are more singular, with higher fuel loads, thus resulting in a more regularly-shaped distribution of highly severe burning patches. A detailed understanding of burn severity patterns and driving factors in a landscape can help develop sustainable forest management and restoration strategies. Practically, fire managers should conduct mechanical fuel treatments or thinning of forests at high-elevation areas to reduce the potential of severe fire behavior and the continuity of fire spread.

Morphological processes and potential for lacustrine delta restoration

Most lacustrine deltas are nowadays highly anthropized systems. River training works during past centuries considered rivers mainly as vectors of water and sediments from a point source to a sink. However, numerous problems have been identified by residents and the competent authorities, such as lakeshore erosion near river deltas, land-use conflicts, and loss of biodiversity without efficiently managing flood risks. These human-induced changes on the deltaic environment led to the deterioration of ecosystem services. This paper focuses on the lacustrine deltas as a complex biotope at the interface of aquatic and terrestrial environments. It proposes a restoration strategy based on two simple geometrical parameters. This methodological approach defines the potential to recover a quasi-natural state of the delta using the slope of the upstream water course and the opening angle of the delta. The result of this work is the creation of a diagram like the ones classifying rivers according to their morphology. This chart will allow the design engineer to determine, given the actual slope of the water course, which shape a delta would ideally adopt without accounting for the downstream lacustrine morphogenic parameters. It will also help in determining the space required for the watercourse to fulfill its natural functions. This methodological approach is applied to a case study: The Reuss Delta, where a restoration project took place. This approach can be used as a pre-study to determine, according to the available space, if it will be possible to improve the morphological dynamics of the delta and to recover its non-anthropized state.

A global assessment of lake restoration in practice: New insights and future perspectives

A global survey of 179 restoration practitioners spanning 65 countries identified the extent of stakeholder engagement as a key factor determining the success or failure of restoration projects. Lack of support across sectors and for funding, policy, monitoring, governance and knowledge assessment of pressures and their effects were most frequently cited as factors contributing to restoration failure.The responses indicate that, although nutrient enrichment is perceived to be the primary issue for lakes globally, the impacts of climate change, hydrological modifications and invasive species are widely recognized as pervasive anthropogenic pressures of global importance. Practitioners recognized that the ecosystem services most impacted by these pressures were recreation and tourism, although in low income countries the provisioning service, aquaculture, was considered most impacted. Ecology-based and/or pressure-related restoration targets had been set for most restoration programs in our survey. However, the strength of the evidence underpinning these targets was often weak and the effects of climate change were rarely considered when setting targets.The most effective and widely used restoration measures target nutrient loading (both catchment and in-lake) while hydrological modifications and the implementation of nature-based solutions are used to a lesser extent. Measures for the control of non-native invasive species are rarely applied and are viewed as being largely ineffective.The results of the survey provide direction for future work. New and emerging pressures, singly and in combination, may require new approaches to lake restoration: for both setting restoration targets and devising restoration strategies. The future of lake restoration depends on joined-up thinking that better integrates science into policy and practice and, most importantly, ensures strong and inclusive stakeholder engagement and collaboration across multiple sectors.

Resilience enhancement of distribution networks based on demand response under extreme scenarios

AbstractIn the context of rare but high‐impact extreme scenarios, such as natural disasters, it is crucial to utilize all available resources, including microgrids and distributed power sources, to restore critical loads in the distribution network as much as possible. This paper proposes a two‐stage resilience enhancement strategy for distribution networks considering post‐disaster topology reconstruction and demand response, aiming to facilitate the recovery of critical loads after disasters. In the first stage, a heuristic algorithm is introduced to determine the post‐disaster topology of the distribution network. In the second stage, by employing a step‐wise elastic load curve, user demand response is incorporated to maximize the socio‐economic value of the restoration, and resilience metrics are computed. Finally, the effectiveness of the proposed resilience restoration strategy is validated through simulations on Case33 and Case69 systems.

Restoring historical grasslands in a desert national park: Resilience or unrecoverable states in an emerging climate?

Ecosystems that supported multiple biological communities during the Holocene and that form alternative stable states pose challenges for identifying appropriate and feasible target communities for ecological restoration and conservation strategies for climate adaptation. In the Chihuahuan Desert, shrubland and perennial grassland alternatively dominated for millennia. However, grassland declined beginning 100–150 years ago during concomitant change in both climate and anthropogenic land use, making it unclear whether restoring grassland is appropriate or feasible in the emerging climate. In Guadalupe Mountains National Park, Texas, USA, we characterized remnant grasslands and performed experiments including outplanting and shrub reduction to assess feasibility and techniques for restoring desert grassland. In remnant patches, grass cover was more temporally variable than shrub cover, as grass cover varied 1.5× between years. Survival of outplanted grasses, either under Larrea tridentata shrubs or in the open with or without woody material encircling outplants for protection, was high (95 %) after seven months but declined to 2 % after 19 months. Response of perennial grass cover was variable after shrub and litter removal on soil surfaces varying in stoniness. Results suggest that 1) restoration can increase native desert grasses but gains may be intermittent; 2) reducing shrubs to facilitate natural expansion of remnant grasses and propagules may be more effective than outplanting grasses; and 3) restoring native grassland by reducing shrubs appears feasible even in dry conditions projected for the emerging climate. Findings that native grasses can increase at least intermittently suggest that practitioners have multiple candidate communities for conservation management and climate change adaptation as grassland, shrubland, and mixed communities each provide unique habitat.

Comprehensive Phenotypic Assessment of Rice Diseases in Cultivated Farms within Okpuitumo Community, Ikwo Local Government Area, Ebonyi State: Implications for Sustainable Rice Crop Management

Rice stands as a pivotal economic crop in Ebonyi State, particularly flourishing in Ikwo Local Government Area. However, the productivity of this vital crop faces significant challenges owing to pestilence and diseases. This study sought to employ advanced PLANTIX image analysis software to systematically screen rice farms suspected of disease infestations within Okpuitumo Community. The aim was to precisely identify prevalent diseases impacting the region's rice cultivation. Leveraging the iterative capabilities of the PLANTIX smartphone application, images of afflicted rice leaves were meticulously examined to identify specific disease types and quantify their severity. The study meticulously selected four villages—Anumocha, Odeligbo, Ettam, and Ogidiga—with three farms per village and five sampling points per farm, employing rigorous randomization protocols. The findings showcased a hierarchy of prevalent rice diseases within the community, highlighting the severity in the following descending order: Potassium deficiency (40%), Brown spot disease (38.33%), Magnesium deficiency (31.67%), Nitrogen deficiency (26.67%), Rice blast (6.67%), Zinc deficiency (5%), Bacterial blight (3.33%), Green horned caterpillar (1.67%), and Leaf scald (1.67%). Notably, the study underscored soil nutrient depletion and the prevalence of brown spot disease as primary impediments to rice cultivation in the area. Consequently, the study advocates for the implementation of robust soil nutrient restoration strategies alongside the development of brown spot disease-resistant or tolerant rice varieties that harmoniously align with the local agro-ecology. These interventions are imperative to safeguard the burgeoning population and the farming community against exacerbating food security crises and persistent poverty challenges

Community-based plant diversity monitoring of a dense-canopy and species-rich tropical forest using airborne LiDAR data

Tropical forests are widely regarded as the Earth’s most important ecosystems, yet they are severely threatened by anthropogenic disturbances. Rapid and extensive monitoring of forest structure and biodiversity is crucial for developing ecologically sound conservation and restoration strategies. Airborne light detection and ranging (LiDAR) can effectively monitor three-dimensional forest canopy structures. Traditionally, LiDAR-based plant diversity estimation has relied on individual tree-based and area-based approaches. However, these approaches face significant challenges in tropical forests due to their complex canopy structures and diverse plant compositions. Therefore, by proposing a novel community-based approach, this study aims to examine the relationship between field-derived biodiversity indices and LiDAR-derived canopy structural metrics of plant communities in a species-rich ForestGEO forest dynamic plot in tropical Hong Kong. Our goal is to determine whether canopy structural metrics extracted from airborne LiDAR data can serve as a robust and efficient alternative for expediting plant diversity monitoring in highly dense and diverse tropical forests. Our results indicate that an integration watershed segmentation technique (for automatic patch-scale plant community delineation), LiDAR-derived canopy structural metrics, and machine learning-based random forest regression analysis can provide accurate predictions of community-based species diversity indices. Among various diversity indices, species richness and the Shannon-Wiener index are most accurately estimated using LiDAR-derived metrics. This study reveals that species richness is predominantly influenced by the existence of multi-layered canopy structures, whereas the Shannon-Wiener index is associated with both multi-layered structures and canopy morphologies. Overall, our findings showcase the immense potential of airborne LiDAR data in advancing the monitoring of structure and biodiversity in dense-canopy and species-rich tropical forests in a spatially explicit manner.

The Growth Rate of Cyphastrea Coral Fragment Under Controlled Condition

The role of corals was very important in coastal and marine ecosystems, whether in terms of education, ecology, and socio-economy. Coral growth can be used as an indicators of marine ecosystem health. This study was aimed to observed survival rate and the growth rate of Cyphastrea coral fragments. Cyphastrea coral fragments growth rate was observed in ex-situ sites and was conducted over a period of 3 months. The micro-fragmentation method was applied in this study. Several factors such as water temperature, light levels, and nutrient concentrations were measured and controlled throughout the study. The growth rate of Cyphastrea coral fragments was measured using ImageJ software. Result of this study exhibited that survival rate of Cyphastrea was 100% and the growth rate of Cyphastrea was 0.012 cm2/month. This study provides important information on the factors that affect the growth rate of Cyphastrea coral fragments in ex-situ sites. The result of this study can be used to develop more effective strategies for coral reef restoration.

Human activities significantly impact China’s net primary production variation from 2001 to 2020

Formulating ecological restoration strategies requires accurately quantifying how climate and anthropogenic factors influence net primary production (NPP). A Carnegie-Ames-Stanford approach (CASA) model was applied to estimate China’s terrestrial NPP from 2001 to 2020. We adopted a random forest (RF) method to identify the main driving forces for NPP change in China. Total NPP in China increased noticeably with a 24.91 Tg C/yr rate, as shown in our results. The significantly increased NPP was mainly attributed to human activities (64.29 ± 0.17%), chiefly due to human management and ecological projects (afforestation or other) fostered vegetation growth. The primary drivers of NPP variation varied in different geographic regions. Climate dominated the NPP dynamic in north China (52.38 ± 0.91%), where the main factor that restricted the increase of NPP was precipitation. Human activities strongly impacted the NPP variation in the remaining regions. Human management measures increased NPP in northwest and southwest China. In the northeast, east, and south-central China, the NPP change resulted from land use change, primarily grassland, cropland, and forest change. Collectively, our study expands the understanding of the driving forces of NPP change, informing different strategies for achieving ecological restoration and carbon neutrality.

Transdermal Microneedles Alleviated Rheumatoid Arthritis by Inducing Immune Tolerance via Skin-Resident Antigen Presenting Cells.

Restoring immune tolerance is the ultimate goal for rheumatoid arthritis (RA) treatment. The most reported oral or intravenous injection routes for the immunization of autoantigens cause gastrointestinal side effects, low patient compliance, and unsatisfied immune tolerance induction. Herein, the use of a transdermal microneedle patch is for the first time investigated to codeliver CII peptide autoantigen and rapamycin for reversing immune disorders of RA. The immunized microneedles efficiently recruit antigen-presenting cells particularly Langerhans cells, and induce tolerogenic dendritic cells at the administration skin site. The tolerogenic dendritic cells further homing to lymph nodes to activate systemic Treg cell differentiation, which upregulates the expression of anti-inflammatory mediators while inhibiting the polarization of Th1/2 and Th17 T cell phenotypes and the expression of inflammatory profiles. As a result, the optimized microneedles nearly completely eliminate RA symptoms and inflammatory infiltrations. Furthermore, it is demonstrated that a low dose of rapamycin is crucial for the successful induction of immune tolerance. The results indicate that a rationally designed microneedle patch is a promising strategy for immune balance restoration with increased immune tolerance induction efficiency and patient compliance.

Power supply restoration strategy of distribution system considering load demand response

The fluctuation of new energy output decreases distribution system reliability and increases the risk of a power outage. Based on this, considering the participation of new energy sources, the influence of load participation demand response on the distribution system is analyzed, the load demand response model is established, and the corresponding unit capacity compensation cost model is proposed. Considering the recovery income, operation cost, demand response compensation cost, and risk, a bi-level optimization model is established to solve the maximum recovery net income of the system. The upper layer aims at maximizing the expected net recovery income of the entire distribution system and proposes a distribution system partition scheme that takes into account the participation of new energy sources and mobile sources. The lower layer aims at maximizing the net recovery income of each partition and solves the recovery time and recovery amount of the load in each partition or the response time and response amount of the load participating in demand response under the condition of limited resources. Finally, the effectiveness of the power supply recovery optimization strategy of the distribution system proposed in this paper is verified by an example.

Beyond the Clinic: why new bioactive restorative materials have really changed Paediatric Dentistry.

The aim of this paper is to remind how evidence-based paediatric dentistry should not only consider the best available scientific evidence relating to the patient's oral conditions but it should also consider the patient's needs beyond the clinic. In a child-family-oriented approach, the best restorative treatment for a primary tooth could not be the proper one for the child as a whole. Preservation of dental tissues as much as possible, without involving the pulp, is called minimal intervention dentistry (MID) and it is closely related to "selective caries removal" and "bioactive restorative materials". To preserve the vitality of a tooth as much as possible, many tools are available on the market (i.e. silver diamine fluoride, ozone) and this could play an important role in a "tailored fit treatment plan" mindset. Minimally invasive direct restorative strategies in paediatric dentistry, is a predictable evidence-based-restorative option for the treatment of vital primary teeth with no need of local anaesthesia and absolute isolation. Beyond the clinical aspect, the clinician expertise is crucial to evaluate the required materials and tools, not only to perform a minimal invasive paediatric dentistry in a safe, efficient and child-friendly way, but for the wellness of all parties involved. In this life-related paradigm new bioactive restorative materials are among the main protagonists.

Silvicultural decisions to formulate forest restoration strategies using geospatial approaches

Restoring deforested and degraded forests is mandatory to guarantee the future of tropical forests. In this study, we formulated silvicultural strategies for restoring tropical forest ecosystems using geospatial approaches to detect deforested and degraded forests based on landscape circumstances. Depending on the overlaid map of the current land cover and its land status, we categorized our forest restoration activity map into three main groups: reforestation, forest rehabilitation, and none. Based on the forest vulnerability, as evidenced by slope and erosion risk, we also allocated the ranks of restoration location into three priorities: first, second, and third. We believe that geospatial modeling maps can assist decision-makers in selecting locations and the most important goals for reforestation in order to reduce greenhouse gas (GHG) emissions in land-based sectors. In order to exactly convey the geospatial data of forest at more detailed landscape scales, such as sub-district and village levels, as well as forest management units, these study need to be enhanced using high-resolution satellite or aerial imageries.

Generating Diverse Optimal Road Management Plans in Post-Disaster by Applying Envelope Multi-Objective Deep Reinforcement Learning

The authors used a data-driven reinforcement learning model for the post-disaster rapid recovery of human mobility, considering human-mobility recovery rate, road connectivity, and travel cost as the recovery components, to generate the reward framework. Each component has relative importance with respect to the others. However, if the preference is different from the original one, the optimal policy may not always be identified. This limitation must be addressed to enhance the robustness and generalizability of the proposed deep Q-network model. Therefore, a set of optimal policies were identified over a predetermined preference space, and the underlying importance was evaluated by applying envelope multi-objective reinforcement learning. The agent used in this study could distinguish the importance of each damaged road based on a given relative preference and derive a road-recovery policy suitable for each criterion. Furthermore, the authors provided the guidelines for constructing the optimal road-management plan. Based on the generalized policy network, the government can access diverse restoration strategies and select the most appropriate one depending on the disaster situation.

A hybrid structural sparse model for image restoration

Structural sparse model (SSM) has achieved great success in various image inverse problems. However, most existing methods just use a single norm, i.e., l1- or l2-norm, to constrain sparse coefficients. Unfortunately, the soft thresholding associated with l1-norm inactivates small coefficients, and the Wiener filtering associated with l2-norm produces an over-smooth solution. Both of them are unfavorable for protecting image details. To cope with this problem, in this paper, we propose a novel hybrid structural sparse model (HSSM) for image restoration, which can preserve image details more effectively. Unlike typical methods, the proposed HSSM uses l1- and l2-norm to constrain sparse coefficients simultaneously, and a scaling factor, which indicates the importance of each coefficient, is introduced to adaptively control the degree of sparsity penalty for each coefficient. Applying the proposed HSSM to image restoration, a general HSSM-based image restoration scheme is established, and the restored image, sparse coefficients and scaling factors can be efficiently solved by using alternating minimization. Furthermore, we prove that the solution of HSSM achieves a compromise between the solutions of l2-norm-based SSM and l1-norm-based SSM. Experimental results on image denoising, deblurring, and deblocking demonstrate that the proposed HSSM-based restoration method can not only outperform many state-of-the-art model-based methods in both PSNR and SSIM, but also compete with recent superior deep learning-based methods.