Table 4. Accuracy assessment of multi-temporal greenhouse maps.
Years Overall Accuracy (%) Kappa Coefficient Producers Accuracy (%) Users Accuracy (%)
2000 93.52% 0.90 71.26% 82.65%
2003 95.55% 0.93 88.70% 90.02%
2006 94.82% 0.93 72.89% 87.34%
2010 96.38% 0.94 79.85% 91.57%
2013 92.85% 0.88 83.37% 88.39%
2015 94.02% 0.91 93.69% 95.29%
2019 94.07% 0.91 88.02% 92.47%

Figure a-g show the distribution of greenhouses in Guanzhong Plain in seven years (2000, 2003,

2006, 2010, 2013, 2015, 2019), respectively. Figure h represents the spatio-temporal distribution

change of greenhouses in a selected area. The classification results (Figure a-g) are corresponded

to four evaluation indicator of each year in Table 4. The overall accuracy of each year and the

kappa coefficient is euqal to or greater than 90%, which is expected to meet the requirements for

quantifying greenhouses’ dynamics distribution.

Table 5. F1-score of greenhouse in seven years.
/ 2000 2003 2006 2010 2013 2015 2019
F1-score 0.44 0.89 0.795 0.854 0.858 0.945 0.902

F1-score (Table 5) was calculated for each year based on two classes (greenhouses and non-greenhouses).

The values of F1-score are high in six years (2003, 2006, 2010, 2013 2015, 2019), which indicates that

a good quality of mapping results. The F1-score is low in 2000, depicting a low quality of the mapping

results.