Forth, the condition indicators or derived values can be used as input parameters
or verification datasets to be assimilated into a crop growth model for estimating and
evaluating the crop growth status (Dadhwal and Ray 2000). To assimilate the crop
condition indicators/indices derived from remote sensing into the crop growth
model, two approaches (Bouman 1995; Dadhwal and Ray 2000) are commonly
used – one is to use the remote sensing-derived information as input parameters to
drive the models, and the other is to use the remote sensing-derived information as
model output verification to optimize the model (Maas 1988). The leaf area index
(LAI) is one of the most frequently used remote sensing-derived information to be
assimilated into the crop growth models. It has been used in both approaches – input
parameter (Doraiswamy 2002; Doraiswamy et al. 2003; Fang et al. 2011) and output
verification (Clevers 1997; Hong et al. 2004) – since LAI is the input for crop growth
models while it is also an intermediate output of many crop growth simulation
models.
Fifth, diagnostic modeling or stress modeling is another way to use remote
sensing-derived crop condition indices into estimating and forecasting the crop
yield (Dadhwal and Ray 2000). Flooding and drought are the two extremes of stress
on crop growth that are of excessive water or deficit water. The relative water content
from crop vegetation indices can be used to evaluate the water condition of the crop
(Kogan 1995; Gao 1996; Yang et al. 2011b). Extreme events, e.g., flooding, may
post different degrees of damages to crops depending on their growth stages, and
their impact may be assessed with remote sensing-derived crop condition indices and
their profiles over the growing season (Di et al. 2013; Shrestha et al. 2013, 2017).
6. Validation: Ground survey at the same time of satellite observations for validation
is the straightforward and accurate approach to verify the results (CHU et al.
2016). However, due to the constraints of resources, ground survey is limited in
the number of samples and frequency of revisits. Alternative approaches to
evaluate the crop condition monitoring are evaluations at aggregated (or coarse)
temporal resolution and/or spatial resolution (Kim and Kaluarachchi 2015; Zhang
et al. 2016), evaluation with higher-resolution satellite or aerial imageries (Huang
et al. 2012; Kim and Kaluarachchi 2015), and evaluation with results from other
methods/models (e.g., statistical survey (Wall et al. 2008; Becker-Reshef et al.
2010b), alternative operational systems (Baruth et al. 2008; Atkinson et al. 2012;
Mladenova et al. 2017), and social outsourcing (Fritz et al. 2012)).
10.3.3 Case Study – Operational Remote Sensing Crop
Condition Monitoring
10.3.3.1 National Crop Progress Monitoring System
The National Crop Progress Monitoring System (NCPMS) is a collaborative project
of George Mason University, USDA NASS, and NASA that developed the operational national crop conditional monitoring system of the contiguous United States
10 Crop Pattern and Status Monitoring
191
or verification datasets to be assimilated into a crop growth model for estimating and
evaluating the crop growth status (Dadhwal and Ray 2000). To assimilate the crop
condition indicators/indices derived from remote sensing into the crop growth
model, two approaches (Bouman 1995; Dadhwal and Ray 2000) are commonly
used – one is to use the remote sensing-derived information as input parameters to
drive the models, and the other is to use the remote sensing-derived information as
model output verification to optimize the model (Maas 1988). The leaf area index
(LAI) is one of the most frequently used remote sensing-derived information to be
assimilated into the crop growth models. It has been used in both approaches – input
parameter (Doraiswamy 2002; Doraiswamy et al. 2003; Fang et al. 2011) and output
verification (Clevers 1997; Hong et al. 2004) – since LAI is the input for crop growth
models while it is also an intermediate output of many crop growth simulation
models.
Fifth, diagnostic modeling or stress modeling is another way to use remote
sensing-derived crop condition indices into estimating and forecasting the crop
yield (Dadhwal and Ray 2000). Flooding and drought are the two extremes of stress
on crop growth that are of excessive water or deficit water. The relative water content
from crop vegetation indices can be used to evaluate the water condition of the crop
(Kogan 1995; Gao 1996; Yang et al. 2011b). Extreme events, e.g., flooding, may
post different degrees of damages to crops depending on their growth stages, and
their impact may be assessed with remote sensing-derived crop condition indices and
their profiles over the growing season (Di et al. 2013; Shrestha et al. 2013, 2017).
6. Validation: Ground survey at the same time of satellite observations for validation
is the straightforward and accurate approach to verify the results (CHU et al.
2016). However, due to the constraints of resources, ground survey is limited in
the number of samples and frequency of revisits. Alternative approaches to
evaluate the crop condition monitoring are evaluations at aggregated (or coarse)
temporal resolution and/or spatial resolution (Kim and Kaluarachchi 2015; Zhang
et al. 2016), evaluation with higher-resolution satellite or aerial imageries (Huang
et al. 2012; Kim and Kaluarachchi 2015), and evaluation with results from other
methods/models (e.g., statistical survey (Wall et al. 2008; Becker-Reshef et al.
2010b), alternative operational systems (Baruth et al. 2008; Atkinson et al. 2012;
Mladenova et al. 2017), and social outsourcing (Fritz et al. 2012)).
10.3.3 Case Study – Operational Remote Sensing Crop
Condition Monitoring
10.3.3.1 National Crop Progress Monitoring System
The National Crop Progress Monitoring System (NCPMS) is a collaborative project
of George Mason University, USDA NASS, and NASA that developed the operational national crop conditional monitoring system of the contiguous United States
10 Crop Pattern and Status Monitoring
191