Comparison of kriging interpolation precision between grid sampling scheme and simple random sampling scheme for precision agriculture

Jiang HouLong, Daibin Wang, Xu Chen, Shu-Duan Liu, Hongfeng Wang, Chao Yang, Li NaJia, Yiyin Chen, Lina GengEURASIAN 2016 · DOI

Abstract

Sampling methods are critical factors that can limit the accuracy of spatial distribution predictions in soil science. This study selected a 10-hectare tobacco-planted field to compare the precision of ordinary kriging and cross-validation methods between grid sampling and simple random sampling (SRS) schemes. Soil samples (0-20 cm) were collected in March 2012, with 115 points for each scheme. Accuracy was evaluated using 36 validation samples. Results indicated that grid sampling provided lower interpolation errors for organic matter (OM), cation exchange capacity (CEC), total nitrogen (TN), and potassium (AK, TK), while SRS performed better for pH and phosphorus (TP, AP). Given the importance of minerals for tobacco cultivation, the grid-sampling scheme is recommended as a practical and cost-effective method for mapping soil properties in tobacco fields.

Introduction

In the era of precision agriculture, uniform fertilizer application is being replaced by variable-rate technology to optimize yields and minimize environmental impact. This transition depends heavily on accurate maps of soil nutrient distribution. However, soil sampling is a significant "bottleneck" due to the high costs and labor involved. Agronomists face a fundamental choice: should samples be collected in a systematic grid or via a simple random approach? This study evaluates which method provides more reliable data for kriging interpolation in a tobacco-growing context.

Methods

The research was conducted on a 10-hectare tobacco field in Chongqing, China. Two sampling strategies were compared, each utilizing 115 points:

  1. Grid Sampling (a): Points were arranged on a systematic 32m x 32m grid.
  2. Simple Random Sampling (b): Points were distributed randomly with spacing between 16m and 32m.
Comparison of simple random sampling (b) and grid sampling (a) schemes across the 10-ha study area.

Nine soil properties were analyzed: pH, organic matter (OM), cation exchange capacity (CEC), total nitrogen (TN), nitrate-nitrogen ($NO_3-N$), total phosphorus (TP), available phosphorus (AP), total potassium (TK), and available potassium (AK). Ordinary Kriging was used to generate continuous spatial maps, and accuracy was verified against 36 independent validation points.

Key Results and Discussion

The study found that most soil properties exhibited moderate to strong spatial dependence. Soil pH and nitrate-N ($NO_3-N$) showed low variation, while all other properties exhibited medium variation. Different mathematical models were required for different nutrients: pH, OM, TN, CEC, TP, and AP fit a spherical model, while $NO_3-N$, TK, and AK fit an exponential model.

Spatial distribution maps of soil properties generated via ordinary kriging for both sampling schemes (a: grid, b: SRS).

The performance of the sampling schemes varied by nutrient:

  • Grid Sampling Superiority: This method yielded lower interpolation errors for OM, CEC, TN, AK, and TK.
  • SRS Superiority: This method performed slightly better for pH, TP, and AP.
  • No Significant Difference: Both methods performed similarly for Nitrate-Nitrogen ($NO_3-N$).

For tobacco production, potassium (AK and TK) is a critical nutrient affecting leaf quality and safety. Since grid sampling provided more accurate maps for these essential minerals, it emerged as the superior choice for this specific agricultural application.

Conclusion

The grid-sampling scheme proved to be more effective for mapping the soil properties most vital to tobacco yield and quality. While the optimal sampling scheme may depend on the specific "noise-to-signal" ratio (nugget-to-sill ratio) of the spatial data, grid sampling offers a practical and cost-effective solution for precision tobacco farming. Future research is needed to verify if these findings hold true across different scales and more varied topographical conditions.

Summary updated

Abstract

Sampling methods are important factors that can potentially limit the accuracy of predictions of spatial distribution patterns. A 10 ha tobacco-planted field was selected to compared the accuracy in predicting the spatial distribution of soil properties by using ordinary kriging and cross validation methods between grid sampling and simple random sampling scheme (SRS). To achieve this objective, we collected soil samples from the topsoil (0-20 cm) in March 2012. Sample numbers of grid sampling and SRS were both 115 points each. Accuracies of spatial interpolation using the two sampling schemes were then evaluated based on validation samples (36 points) and deviations of the estimates. The results suggested that soil pH and nitrate-N (NO3-N) had low variation, whereas all other soil properties exhibited medium variation. Soil pH, organic matter (OM), total nitrogen (TN), cation exchange capacity (CEC), total phosphorus (TP) and available phosphorus (AP) matched the spherical model, whereas the remaining variables fit an exponential model with both sampling methods. The interpolation error of soil pH, TP, and AP was the lowest in SRS. The errors of interpolation for OM, CEC, TN, available potassium (AK) and total potassium (TK) were the lowest for grid sampling. The interpolation precisions of the soil NO3-N showed no significant differences between the two sampling schemes. Considering our data on interpolation precision and the importance of minerals for cultivation of flue-cured tobacco, the grid-sampling scheme should be used in tobacco-planted fields to determine the spatial distribution of soil properties. The grid-sampling method can be applied in a practical and cost-effective manner to facilitate soil sampling in tobacco-planted field.
BibTeX
@inproceedings{houlong2016comparison,
  author = {Jiang HouLong and Daibin Wang and Xu Chen and Shu-Duan Liu and Hongfeng Wang and Chao Yang and Li NaJia and Yiyin Chen and Lina Geng},
  title = {{Comparison of kriging interpolation precision between grid sampling scheme and simple random sampling scheme for precision agriculture}},
  booktitle = {EURASIAN},
  year = {2016},
  doi = {10.18393/ejss.2016.1.062-073},
}
Figure