Mapping Forest Degradation and Contributing Factors in a Tropical Dry Forest

Abstract

Forest degradation reduces biomass density, contributes to greenhouse gas emissions, and affects biodiversity and natural resources available for local communities. Previous studies have reported that gross emissions from forest degradation might be higher than from deforestation, due to the larger area affected by the first process. The quantification of forest degradation with remote sensing has large uncertainty, mainly because the subtle and gradual changes in forest are challenging to detect, and sometimes these changes happen below the canopy cover which the optical sensors cannot see. The objective of this work is to map the degraded forests and the most relevant biophysical and socio-economic factors contributing to such degradation in the dry tropics. We mapped the degraded forests by modeling forest biophysical parameters with multi-temporal optical data of Landsat-8 and Sentinel-2 and identified the most relevant biophysical and socio-economic factors that can be associated with forest degradation. We included three biophysical variables and 11 socio-economic variables including parceled land and land in ejido property and used multiple linear regression to relate those variables with identified degraded forests. We identified 62,878 ha of tropical dry forest in a degraded state, cover 49.91% of the forest area. The most relevant biophysical factor was distance to settlements and the most relevant socio-economic factor was percentage of parceled land property (private land). Both factors were negatively associated with the mapped degraded forests. Since parceled land and land in ejido property are strongly and positively correlated (Pearson’s r = 0.82, p < 0.001), it suggests that ejido property, as a form of land tenure, plays an important role in preventing forest degradation. This experiment presents a possible way to measure and understand degradation which may help finding solutions to slow down forest degradation and promote forest restoration.

An Analysis of the Factors Affecting Forest Mortality and Research on Forecasting Models in Southern China: A Case Study in Zhejiang Province

“…We employed a selection of traditional Linear Regression, AIC Stepwise Regression, Ridge Regression, and Lasso Regression models to screen the predictors. Linear Regression primarily seeks to minimize the error between the actual value and the predicted value through the least squares method to identify the best predictors [32][33][34]. Stepwise Regression aims to maximize the predictive power of the model by identifying the optimal feature subset from a given set for feature selection [35,36].…”

Prediction and mapping of land degradation in the Batanghari watershed, Sumatra, Indonesia: utilizing multi-source geospatial data and machine learning modeling techniques

“…Meanwhile, the mTPI and Landforms data obtained from Conservation Sciences Partners have the spatial resolution of 270 m and 90 m, respectively(Theobald et al 2015). (b) Biophysical variablesBiophysical is one of the important and relevant variables that have a contribution related to the phenomenon of land degradation in an environment(Lestariningsih et al 2018;Jiménez-Rodríguez et al 2022). In addition, monitoring of biophysical variables over a long period of time is very important to overcome challenges regarding productivity, climate change impacts and food security in a region(Kandasamy 2014…”