Rapid Crop Cover Mapping for the Conterminous United States.
Sci Rep
; 8(1): 8631, 2018 06 05.
Article
em En
| MEDLINE
| ID: mdl-29872107
ABSTRACT
Timely crop cover maps with sufficient resolution are important components to various environmental planning and research applications. Through the modification and use of a previously developed crop classification model (CCM), which was originally developed to generate historical annual crop cover maps, we hypothesized that such crop cover maps could be generated rapidly during the growing season. Through a process of incrementally removing weekly and monthly independent variables from the CCM and implementing a 'two model mapping' approach, we found it viable to generate conterminous United States-wide rapid crop cover maps at a resolution of 250 m for the current year by the month of September. In this approach, we divided the CCM model into one 'crop type model' to handle the classification of nine specific crops and a second, binary model to classify the presence or absence of 'other' crops. Under the two model mapping approach, the training errors were 0.8% and 1.5% for the crop type and binary model, respectively, while test errors were 5.5% and 6.4%, respectively. With spatial mapping accuracies for annual maps reaching upwards of 70%, this approach demonstrated a strong potential for generating rapid crop cover maps by the 1st of September.
Texto completo:
1
Base de dados:
MEDLINE
Tipo de estudo:
Prognostic_studies
Idioma:
En
Revista:
Sci Rep
Ano de publicação:
2018
Tipo de documento:
Article
País de afiliação:
Estados Unidos