School of Biological and Physical Sciences

Adaptive moment estimation of the spatial autoregressive model parameters

Thu, 01 Jan 2026 00:00:00 GMT

Adaptive moment estimation of the spatial autoregressive model parameters Otieno, Michael Oduor; Wagala, Adolphus; Kerich, Gregory Traditionally, estimation of the spatial autoregressive (SAR) model parameters has been done using the maximum likelihood estimation (MLE) technique. Several variants of the MLE technique, such as quasi-MLE (QMLE) and adjusted-QMLE, among others, have also been used. However, in the presence of large data sets, MLE has been found to be computationally costly due to the matrix inversion involved in maximizing the log-likelihood function. The adaptive moment (ADAM) estimation method is a viable alternative. It eliminates the cost of matrix inversion in the widely used MLE method. Moreover, it eliminates the need to store the whole data set since the gradients can be computed from every data point, or a batch of data points sampled from the whole data set. However, ADAM has been developed and widely used for independent data. This study seeks to design ADAM for use in a spatial data setting. ADAM is found to perform better than the standard MLE in large data setting since it has the least values of Akaike Information Criterion and higher values of the McFadden’s R-squared. It also converges faster than MLE and hence is most appropriate for large data sets.

Assessment of water quality by physico-chemical properties and nematodes as bio- indicators along river Sosiani in Uasin Gishu, Kenya

Thu, 01 Jan 2026 00:00:00 GMT

Assessment of water quality by physico-chemical properties and nematodes as bio- indicators along river Sosiani in Uasin Gishu, Kenya Kemboi, Phanuel Kimurgor Water quality in freshwater systems is increasingly threatened by industrial, domestic, and agricultural pollution, which introduces contaminants that degrade ecological integrity and pose risks to human health. River Sosiani is a critical water resource for domestic, agricultural, and industrial use; however, its quality has deteriorated due to increasing anthropogenic activities, necessitating more reliable and integrative monitoring approaches. Conventional chemical assessments provide only short-term snapshots of water conditions and fail to capture cumulative or long-term pollution effects. In contrast, biological monitoring using nematodes as bioindicators provides a more comprehensive assessment of both current and historical water quality. The general objective of this study was to apply nematodes as bioindicators for monitoring pollution in River Sosiani. Specifically, the study aimed to isolate and quantify nematodes from water and sediment samples and to examine the relationship between nematode community characteristics and pollution levels. The methodology involved collecting water and sediment samples from four key locations along River Sosiani: Kaptagat (control site), Kipkorgot, Kisumu Bridge, and Huruma. Nematodes were extracted from both water and sediment using a modified Baermann funnel technique. The densities of the nematodes were calculated as population densities (PD), expressed as the number of individuals per unit volume of water or per unit weight of sediment. The relationship between nematodes and pollution levels was assessed by correlating nematode population densities with physicochemical parameters, including dissolved oxygen and turbidity (as well as pH, temperature, and electrical conductivity). Results indicated a clear spatial trend in nematode abundance, with low nematode numbers in Kaptagat (93 individuals) and Kipkorgot (167), and significantly higher numbers in Kisumu Bridge (1344) and Huruma (1792), indicating a progressive increase along the pollution gradient, (P = 0.005). These trends corresponded with changes in physicochemical parameters, where pH decreased slightly from 7.39 at Kaptagat to 6.82 at Huruma, turbidity increased from 3.50 NTU to 8.36 NTU at Huruma, dissolved oxygen decreased from 5.45 mg/L to 3.55 mg/L at Huruma, temperature increased from 13.30 °C to 18.93°C at Huruma, and electrical to 18.93 °C to 18.93°C at Huruma, and electrical at Huruma, and electrical conductivity increased from 48.71 S/m to 174.35 S/m at Huruma, demonstrating a strong relationship between deteriorating water quality and increased nematode abundance, suggesting dominance of pollution-tolerant species. In conclusion, nematode-based assessments provide a reliable and integrative tool for assessing water quality and ecological changes. The study recommends their use alongside conventional chemical analyses for enhanced monitoring and sustainable management of River Sosiani and similar ecosystems.

Analysis of climate change impacts on plant biodiversity and livelihoods among Maasai women in Narok County, Kenya

Wed, 01 Jan 2025 00:00:00 GMT

Analysis of climate change impacts on plant biodiversity and livelihoods among Maasai women in Narok County, Kenya Sinteria, John Kishoyian Climate change poses critical threats to plant biodiversity and pastoral livelihoods in sub-Saharan Africa, yet knowledge gaps persist regarding gender-differentiated impacts of climate-biodiversity interactions on women's livelihood vulnerability. This study assessed climate change impacts on plant species diversity and implications for Maasai women's livelihoods in Narok County, Kenya. A mixed-methods convergent design was employed, integrating quantitative climate data analysis (1990-2020), systematic botanical surveys, structured questionnaires (n=100 Maasai women), and qualitative assessments through focus group discussions (n=24 groups) and key informant interviews (n=15 traditional experts). Climate data from Kenya Meteorological Department and NASA POWER database were analyzed using Mann- Kendall trend tests and Sen's slope estimators. Ethnobotanical surveys utilized systematic transect-quadrat sampling across eight locations. Vulnerability assessment employed Hahn's Livelihood Vulnerability Index framework. Results revealed statistically significant warming of 0.35°C per decade (Mann-Kendall τ=0.312, p<0.01) with extreme temperature events reaching 2.35°C above baseline. Precipitation showed high inter-annual variability (coefficient of variation=31.2%) with significant seasonal shifts including September increases (τ=0.338, p=0.009) and February decreases approaching significance (τ=-0.251, p=0.054). Botanical surveys documented 89 plant species across 33 families, with medicinal uses dominating (36% of species), followed by construction materials (13%) and fodder (11%). Diversity indices indicated moderate levels (Shannon-Weiner H'=1.335; Simpson's D=0.421). Critical conservation concerns emerged with 31 species (35%) occurring in single locations and 25 species at critically low densities, indicating high extinction risk. The Climate Vulnerability Index (4.4) demonstrated moderate vulnerability, with strong adaptive capacity (10.4) buffering high plant-based sensitivity (3.8) and moderate climate exposure (2.2). Climate awareness was exceptionally high (91% of respondents), with strong correspondence between women's perceptions and meteorological data validating traditional ecological knowledge systems. The study conclusively demonstrates that climate change significantly impacts plant biodiversity with direct implications for Maasai women's livelihoods. Despite strong traditional knowledge and social capital through cooperatives, communities face climate risks and biodiversity loss that threaten healthcare access, food security, and cultural practices. Key recommendations include establishing community conservancies with women as primary managers, implementing climate-smart plant management integrating traditional and scientific knowledge, strengthening women's cooperatives for economic resilience, developing integrated climate information systems, and creating intergenerational knowledge transfer programs. These findings advance understanding of the climate-biodiversity-gender nexus and inform evidence-based policy interventions for pastoral communities navigating climate uncertainty.

Mathematical modeling of energy mix and optimization of renewable resources

Wed, 01 Jan 2025 00:00:00 GMT

Mathematical modeling of energy mix and optimization of renewable resources Sigei, Kipkirui Robert Energy, as both a direct and indirect fundamental life-supporting resource, has experienced a steady rise in domestic and industrial demand, driven by technological advancement, population growth, and economic expansion. Various sources of energy including fossil fuels, hydroelectric power, geothermal energy, wind, solar, and nuclear are available in different proportions, each with distinct cost structures and environmental impacts. The challenge of meeting these diverse needs while minimizing production and distribution costs, conserving the environment, and reducing wastage has evolved into a complex multi-objective problem. This research focuses on the mathematical modelling of the optimal energy mix and the optimization of renewable resources, with particular emphasis on individualized demand profiles. The objectives are threefold: first, to formulate a mathematical model for analysing the dynamics of energy demand, production, and distribution; second, to determine the parameter thresholds that guarantee stability and robustness of the optimal energy mix; and third, to develop a smart grid feedback model using adaptive neural networks capable of automatically maintaining the desired energy balance. The methodology entails formulating a system of differential equations to represent the energy system, expressing it in state-space form, and applying Laplace transforms to derive transfer functions. These will be analysed for sensitivity, stability, and robustness using Nyquist and Bode plot criteria. MATLAB–Simulink, equipped with neural network modules, will then be employed to simulate and implement an intelligent, adaptive feedback control system. Through these simulations, the study will integrate real-time learning and self-adjustment capabilities to align production with demand in the most efficient manner. The anticipated outcome is an automated, smart distribution system capable of dynamically meeting individualized energy requirements at the lowest possible cost, while enhancing the utilization of renewable sources and reducing reliance on non-renewable options. Ultimately, this approach aims to promote environmental sustainability through increased adoption of green energy technologies.