Table 3 Presents the model testing performance metrics of the PMC and LMC of winter wheat with only vegetation indices as input.

From: Novel indices and multi-source data fusion for monitoring plant moisture stress in winter wheat fields

Moisture type

Index type

RF

PLSR

SVM

ANN

R2

RMSE

MAE

R2

RMSE

MAE

R2

RMSE

MAE

R2

RMSE

MAE

LMC

RSI

0.756

3.422

2.288

0.613

4.315

3.408

0.824

2.910

2.111

0.727

3.621

2.745

NDSI

0.639

4.165

2.701

0.597

4.403

3.750

0.732

3.587

2.542

0.488

4.963

3.806

Published indices

0.752

3.454

2.298

-3.105

14.050

5.498

0.726

3.627

2.580

0.578

4.505

2.905

PMC

RSI

0.817

3.017

2.220

0.807

3.100

2.559

0.679

3.993

2.628

0.849

2.736

2.131

NDSI

0.810

3.076

2.268

0.825

2.950

2.472

0.828

2.927

2.391

0.844

2.784

2.284

Published indices

0.809

3.084

2.136

0.831

2.895

2.313

0.720

3.729

2.448

0.828

2.924

2.427

  1. RF is the random forest model, PLSR is the partial least squares regression model, SVM is the support vector machine model and ANN is the artificial neural network model. PMC and LMC are plant moisture content and leaf moisture content, respectively.
Back to article page