Elaboración de una base de datos de herbazal de páramo en la zona centro norte del Ecuador mediante herramientas geobi.Alexis Gonzalo Cifuentes Moyaa
The objective of this work was the development of a moorland herbaceous database in the north central zone of Ecuador. SIGs were used to achieve the objective, as well as the GEOBIA methodology proposed by Urbanski in 2013, and LandSat 8 images of the year 2016. This methodology was modified to incr...
Spremljeno u:
| Glavni autor: | |
|---|---|
| Format: | bachelorThesis |
| Jezik: | spa |
| Izdano: |
2018
|
| Teme: | |
| Online pristup: | http://dspace.unach.edu.ec/handle/51000/5172 |
| Oznake: |
Dodaj oznaku
Bez oznaka, Budi prvi tko označuje ovaj zapis!
|
| Sažetak: | The objective of this work was the development of a moorland herbaceous database in the north central zone of Ecuador. SIGs were used to achieve the objective, as well as the GEOBIA methodology proposed by Urbanski in 2013, and LandSat 8 images of the year 2016. This methodology was modified to increase the classification accuracy, starting with the calculation of 17 indexes from models of elevation and spectral bands, the data mining by artificial intelligence allowed to generate 200 decision trees, from which the one with the highest percentage of precision was taken (93%) and with the lowest number of indexes, only 5 (VARIg, NBR2, EVI2, DEM and Slope) that accompanied the reflectance in the classification. Although the MAE reports in 2013 the existence of 770 922.59 hectares of moorland, in the provinces of Bolivar, Tungurahua, Cotopaxi, Napo, Pichincha, Imbabura and Carchi, our study suggest that in this moorland there are 359 584.87 hectares which correspond to the “Herbazal” ecosystem (46.64%), 246 070.33 hectares correspond to intervened moorland (31.92%), 117 950.21 hectares belong to Evergreen Forest (15.30%), 46 633.41 hectares have characteristics of Water Bodies (6.05%) and 683.57 hectares have a snow layer (0.09%). In this way, our results allow us to conclude that a continuous monitoring of this ecosystem is necessary since the total intervention already exceeds 30%. The methodology developed in this study is very convenient for a rapid, efficient and economic monitoring of ecosystems. |
|---|