Project - Building a model for simulation of the "Dose-Response" between pesticide application and disease spread on Pepper crop cultivation to support for the development of solutions for cleaner farming practices in agriculture
The overuse of chemical fertilizers and pesticides in agricultural activities in Vietnam has caused the degradation of the soil environment, causing negative impacts on human health and other environmental factors.
This study aims to develop models capable of predicting the health and growth of pepper plants under different strategies of fungicides application, and to analyze the residues of these chemical products in the soil.
Data used to develop the models and estimate model parameters were obtained through field surveys and measurements along with the implementation of a greenhouse experiment to examine the “Dose – Response” model with more than 600 young pepper plants divided into 28 treatments. Bayesian statistical models were used to analyze these experimental data to detect the effects of fungicides on plant health and mortality.
GIS and remote sensing techniques combined with multi-criteria analysis methods and fuzzy logic were also applied to develop a decision support system for the sustainable management of the soil environment in farming activities.
Field surveys and measurements
Greenhouse experiment
Results
PPR2021- an application to predict temporal and spatial growth response of pepper plants under fungicide treatments and soil conditions using Random Forest Algorithm
Land suitability zoning map for pepper cultivation using fuzzy logic
Vegetation classification
Ecosystem service zoning map – 2015
Ecosystem service zoning map – 2019
Simulation of fungicide fate in soil and response of plant health
Frequentist and Bayesian data analysis
A result of fungicide usage of the social survey campaign
A result of data analysis for Alumium residual in the greenhouse experiment
Boxplot illustration for pepper growth data
Marginal distribution of pathogen mortality with Bayesian approach
A parameter estiamtion with MCMC chains
Bayesian model comparision result
Project - Modelling mangrove dynamics under changing environmental condition
Field measurements in Can Gio mangrove forest
Arriving at a sampling location
Marking GPS coordiantes and plant species
Taking a sediment sample at 50cm depth
Moving to sampling location in the forest
Can Gio mangrove forest management and prediction Software
A supportive application that helps managers to store the forestry sampling data and report statistical charts. It has simulation features that are capable of predicting forest health and biomass under multiple scenarios and field conditions. Extensive expansion for GIS application and auto data source recognition are also available in this application.
Modelling
HM concentration in soil in 3 scenarios
Spatial and temporal distribution of HM
Compare simulation biomass with RS results
A simulation result of mangrove biomass from 2018 to 2050