Modeling Crop Phenology in the US Corn Belt Using Spatially Referenced SMOS Satellite Data
- PDF / 1,205,160 Bytes
- 19 Pages / 496.063 x 720 pts Page_size
- 99 Downloads / 169 Views
Modeling Crop Phenology in the US Corn Belt Using Spatially Referenced SMOS Satellite Data Colin Lewis-Beck, Zhengyuan Zhu , Victoria Walker, and Brian Hornbuckle Satellite measurements follow the growth and senescence of vegetation aid in monitoring crop development within and across growing seasons. For example, identifying when crops reach their peak growth stage or modeling the seasonal growing cycle is useful for agronomists and climatologists. In this paper, we analyze remote sensing data from an intensively cultivated agricultural region in the Midwest to provide new information about crop phenology. There is both a temporal and spatial dimension to the data as they are collected every 12 – 36 hours over regions approximately the size of a 45 km diameter circle. We represent the measurements using a functional data approach and account for spatial dependence between locations through the functional curve coefficients. Modeling across multiple growing years, and including growing degree days as a covariate, we estimate the timing for when crops reach their peak each season and make predictions at unobserved locations. Supplementary materials accompanying this paper appear online. Key Words: Bayesian estimation; Remote sensing; SMOS; Spatial model; Stan; Vegetation indices.
1. INTRODUCTION The European Space Agency’s Soil Moisture and Ocean Salinity (SMOS) satellite has recently been shown to collect data relevant to farmers and agronomists (Hornbuckle et al. 2016). The new data product, referred to as Level 2 τ in the remote sensing literature, measures the water column density of vegetation, which is proportional to the amount of ground vegetation (Jackson and Schmugge 1991). After plants emerge, the water column density increases, reaching a peak during the reproductive growth stage. As an example, for maize, following the third reproductive stage (R3), the plant begins to lose water, undergo
C. Lewis-Beck, University of Iowa, Iowa City, USA. Z. Zhu (B) · B. Hornbuckle Iowa State University, Ames, USA Z. Zhu (E-mail: [email protected]). V. Walker, University of Montana, Missoula, USA. © 2020 International Biometric Society Journal of Agricultural, Biological, and Environmental Statistics https://doi.org/10.1007/s13253-020-00419-x
C. Lewis-Beck et al.
senescence, and dry out. Previous research has empirically found a connection between Level 2 τ and the stages of crop development (Patton and Hornbuckle 2013; Lawrence et al. 2014). The goal of this paper is to use the SMOS Level 2 τ data product (herein referred to as τ ) to provide new information about crop phenology. There are other data sets that capture similar information as SMOS; for example, the Normalized Difference Vegetation Index (NDVI) and the National Agricultural Statistics Service (NASS) Crop Progress Reports. However, the SMOS satellite data are unique due to its combination of high temporal (observations are collected approximately every 12 to 36 hours) and spatial resolution (over 10 times better than USDA ground-based visual surveys).
Data Loading...
