Non-timber forest products and environmental valuation of National Forest Pacotuba
Name: ELVIS RICARDO FIGUEIRA BRANCO
Publication date: 08/07/2016
Advisor:
Name |
Role |
|---|---|
| ALEXANDRE ROSA DOS SANTOS | Advisor * |
| JOSÉ EDUARDO MACEDO PEZZOPANE | Co-advisor * |
Examining board:
| Name |
Role |
|---|---|
| ALEXANDRE ROSA DOS SANTOS | Advisor * |
| AUREO BANHOS DOS SANTOS | External Examiner * |
| JOSÉ EDUARDO MACEDO PEZZOPANE | Co advisor * |
| RODRIGO SOBREIRA ALEXANDRE | Internal Examiner * |
Summary: The application of remote sensing techniques in time series have become a strong tool that have been highlighted in the scientific community by allow rapid and low-cost evaluation, within an accuracy margin. Thus, the frequent recording of images by satellite sensors covering large areas of the Earth surface allows the construction and analysis of time series of vegetation data of different physiognomy, assisting in the dynamic study of vegetation and the spatial arrangement of different intensities drought event. The current study aimed to analyze drought occurrences and time-spatial trends in vegetation and to link them to climate change from 2007 to 2015 in the Sooretama biological reserve and surroundings. For this, were used NDVI, EVI and LST images of MODIS sensor to time-spatial analysis of droughts, was used the Vegetation Condition Index (VCI). Was calculated the Pearson's correlation between the mean values of vegetation index and climate variables in order to determine the most appropriate index for the study area and subsequent employment in the drought index VCI, used for monitoring drought. The drought occurrence images and graphics in severe and extreme classes were compared with the rainfall accumulated and its anomaly and water deficit accumulated and the anomaly for each season. In addition, was obtained the average, maximum and minimum values of VCI and related to the average, maximum and minimum values of LST through graphical analysis and regression. Then, was calculated the LST anomaly and crossed with the seasons that had higher drought extensions of greater severity. The analysis of the inter-annual trends of the time series of vegetation indexes were made through Mann-Kendal monotonic trend and seasonal trends analysis methodologies. The images were imported in .img format of TerrSet software, in wich was used the Earth Trends modeler module (ETM) for the trends analysis and processing of the behavior of vegetation indexes. It was created a time series file for each group of images, NDVI, EVI, in which each series consists of a pair of files: A scan file containing the time series images, in .rgf format and a documentation file that describes the temporal characteristics of the series in .tsf format. The results indicates that the EVI index of MODIS sensor showed higher significant correlations with the meteorological variables and great potential for drought occurrences analysis to regions with high density of biomass as native forests. In the years 2007, 2013 and 2015 occurred extreme and severe order drought in larger extensions compared to the other years. The VCI Index its presented suitable for drought occurrences monitoring in the study area, with temporal and spatial concordance with environmental variables and occurrences of El Niño. Over the period analyzed, it was found that there WHERE a decrease of biomass observed in both vegetation indexes through of the negative trends observed in images time series, being more evident in the EVI. The natural forest areas showed the greatest decrease in vegetative vigor observed in the significance images. The annual average values of EVI and NDVI and its decrease over the years showed agreement with the rainfall decreasing and water stress increasing. The data obtained from the MODIS sensor, NDVI, EVI and LST, proved suitable to temporal-spatio analysis of drought occurrences and vegetation trends of the study area.
