Asutosh College (Bengali: আশুতোষ কলেজ) is a college affiliated to the University of Calcutta, situated in Southern Kolkata, close to the Jatin Das Park Metro Station, gate No. 2 . It was established in 1916 as the South Suburban College, under the stewardship of educationist Sir Ashutosh Mukherjee, who was the then vice-chancellor of the University of Calcutta. After the death of Sir Mukherjee, the college was renamed as Asutosh College in 1924. The principal of the college is Dr. Dipak Kumar Kar. It is affiliated to the Calcutta University .The first college in West Bengal to be accredited by the National Assessment and Accreditation Council in 2002, it was given an A grade with a CGPA score of 3.22 in 2016, helping it secure its position among the top four affiliated colleges of West Bengal.
In this work, we constrain the parameter space of the Ricci-Cubic Holographic Dark Energy (RCHDE) model using several observational datasets, including Hubble parameter measurements, cosmic chronometer (CC) data, Baryon Acoustic Oscillation (BAO) data, and recent DESI observations. The RCHDE model is constructed from a cubic curvature invariant formed through cubic contractions of the Ricci and Riemann tensors. To estimate the model parameters, we employ the Markov Chain Monte Carlo (MCMC) sampling technique within a Bayesian inference framework. The resulting likelihood contours provide both marginalized and joint posterior distributions of the model parameters. The best-fit cosmological evolution predicted by the RCHDE model is reconstructed and compared with observational H(z) measurements as well as with the standard ACDM cosmological model. The best-fit value obtained in our model exhibits a moderate Hubble tension of approximately 2.3 sigma with respect to the reference value for ACDM. While this indicates a noticeable discrepancy, it remains significantly lower than the similar to 5 sigma tension typically reported between early-and late-Universe measurements, suggesting a partial alleviation of the tension. In addition to the statistical parameter estimation, we perform an enhanced machine learning analysis using observational Hubble parameter data. Several supervised regression algorithms are implemented to reconstruct the expansion history of the universe and to test the predictive capability of the RCHDE model from a data-driven perspective. Various graphical analyses are presented to illustrate the performance of the machine learning models. The results demonstrate a strong consistency between the predictions of the machine learning models, the theoretical RCHDE model, and the observational data. We have done a comparative stability analysis between different holographic dark energy models using the squared speed of sound, where it is seen that the RCHDE model does not have any upper hand over its counterparts. Finally, the cosmic coincidence problem is tested to compare the efficiency of the RCHDE model in comparison to other models. It is found that the RCHDE model produced a significant alleviation to the cosmic coincidence problem, outshining its counterparts.
In this work, we explore the new agegraphic dark energy model within the framework of Loop Quantum Cosmology (LQC). A quantum gravitational perspective on the dark energy evolution is explored. A combination of cold dark matter and dark energy in the form of New Agegraphic dark energy is considered with LQC as the background gravity theory. Both the interacting and non-interacting scenarios between dark energy and matter are considered. Various cosmological parameters, like the equation of state parameter, deceleration parameter, and statefinder parameters, are studied. The squared speed of sound is investigated to get an idea about the stability of the system. An observational data analysis using recent cosmological data and the Markov chain Monte Carlo algorithm is performed to constrain the free parameter space of the model. Our findings imply that the interaction of loop quantum effects with agegraphic dark energy offers a nonsingular origin scenario and a theoretically sound and observationally compatible explanation of the universe’s late-time acceleration.
Rapid peri-urban expansion in the Global South has intensified land surface temperature (LST) heterogeneity, yet the seasonal drivers and spatial logic of this thermal variability remain poorly understood. This study investigates the seasonal controls of LST in New Town, Rajarhat (NTR), India, by integrating remote sensing, explainable machine learning, and spatial clustering. Seasonal (summer and winter) LST was derived from Landsat imagery and modelled using multiple machine learning algorithms, with LightGBM and CatBoost demonstrating the highest predictive skill (R2 up to 0.99 in summer and 0.96 in winter). SHapley Additive Explanations (SHAP) were employed to quantify the direction and magnitude of individual predictor influences, including road-network proximity, built morphology, surface bareness, vegetation, moisture, and population density. Results reveal a pronounced seasonal shift in thermal controls. Summer LST is primarily governed by proximity to primary and secondary roads and high surface bareness, with areas within 50 m of major roads exhibiting temperature anomalies exceeding +1.5 degrees C due to combined radiative and advective effects. In contrast, winter LST is dominated by surface material properties, with the bare-soil index alone accounting for nearly 80% of model importance, reflecting emissivity and albedo-driven processes under reduced insolation. SHAP-based clustering further identifies distinct peri-urban thermal regimes, highlighting the coexistence of bare-built hot-spots and moisture-mediated cooling zones that vary seasonally. These findings demonstrate that peri-urban thermal environments are process-driven, seasonally adaptive, and spatially heterogeneous rather than density-dependent. The study provides a transferable, explainable framework for season-specific urban heat mitigation and climate-resilient planning in rapidly urbanising regions.
Soil salinization affects more than 1 billion hectares globally and 6.74 million hectares in India, causing substantial productivity losses and degradation of soil ecological function. This review synthesizes evidence from the literature (1977-2025) to evaluate mulberry (Morus spp.) as a nature-based solution (NbS) for bio-reclamation of salinity-degraded landscapes within agroforestry restoration systems. Genotype-level analyses indicate mulberry tolerance to soil electrical conductivity (EC) up to 23 dS m-1, with operational EC thresholds of 4-8, 8-15, and 15-23 dS m-1. Mulberry enhances ion regulation, antioxidant defense and progressive soil structural recovery while supporting sericulture-linked livelihoods. An artificial intelligence (AI)-enabled decision-support system achieved strong predictive performance (coefficient of determination [R2] = 0.857; root-mean-square error [RMSE] = 0.769), illustrating how integration of ecological mechanisms and data-driven modeling can enable scalable, climate-resilient land restoration planning.
Urban expansion systematically alters land use, producing cumulative impacts on surface-level environmental quality. Addressing these processes requires a computational framework capable of observing and modeling environmental dynamics within a unified urban system. This study formulates a computational framework to model trajectories of surface environmental quality in a metropolitan region, using the Kolkata Urban Agglomeration as a representative case. Multi-temporal Landsat data (1991–2021) were employed to model land-use and land-cover (LULC) transition processes and their relationships with key land surface parameters. Urban environmental quality was quantified using a Land Surface Environmental Quality Index (LSEQI) derived through an aggregate application of multiple multi-criteria decision analysis (MCDA) techniques, including AHP, TOPSIS, EDAS, and VIKOR, followed by ensemble integration to enhance robustness. Model performance was evaluated using statistical consistency measures and proxy validation based on community perception data. To extend the analysis beyond historical assessment, a three-stage predictive architecture was developed, incorporating artificial neural networks for future LULC simulation, dummy regression to quantify LULC–environmental parameter relationships, and MCDA-based synthesis to generate projected LSEQI for 2030. Results indicate a nonlinear urban transformation, with built-up areas expanding from 17.33% in 1991 to 59.80% in 2021, accompanied by continuous losses of vegetative and agricultural land and progressive wetland degradation. Ensemble LSEQI outputs reveal increasing spatial concentration of low environmental quality within dense metropolitan zones. Predictive simulations suggest built-up areas will exceed 68% by 2030, implying further environmental stress. The proposed framework offers transferable computational tools for diagnosing and forecasting environmental consequences of metropolitan growth.
