Author Affiliations
1College of Meteorology and Oceanography, National University of Defense Technology, Changsha 410000, Hunan, China2National Satellite Meteorological Center, China Meteorological Administration, Beijing 100081, Chinashow less
Fig. 1. Flow chart of data matching between FY-4A and CloudSat
Fig. 2. Schemes of CBH retrieval for FY-4A designed in this paper
Fig. 3. Variation of RMSE of eight types of clouds with decision trees number in RF model and GBT model. (a) Variation of RMSE with decision trees number in RF model; (b) variation of RMSE with decision trees number in GBT model
Fig. 4. Variation of RMSE of eight types of clouds with maximum depth of decision trees in RF model and GBT model. (a) Variation of RMSE with maximum depth of decision trees in RF model; (b) variation of RMSE with maximum depth of decision trees in GBT model
Fig. 5. Variation of RMSE of samples on training and validation datasets with decision trees number in RF model and GBT model. (a) Variation of RMSE with decision trees number in RF model; (b) variation of RMSE with decision trees number in GBT model
Fig. 6. Variation of RMSE of samples on training and validation datasets with maximum depth of decision trees in RF model and GBT model. (a) Variation of RMSE with maximum depth of decision trees in RF model; (b) variation of RMSE with maximum depth of decision trees in GBT model
Fig. 7. Retrieval results of models of two schemes on test dataset. (a) Scheme one; (b) scheme two
Fig. 8. Retrieval flow of CBH for FY-4A in practical application
Fig. 9. CBH retrieved from models of two schemes and the comparison with CBH from CloudSat. (a) Cloud types obtained according by the model proposed in Ref. [
32]; (b) CBH retrieved from the cloud types of Fig. 9(a) and the model of scheme one; (c) CBH retrieved from the model of scheme two; (d) comparison between the cloud types of CloudSat and the model proposed in Ref. [
32], and comparison among CBH retrieved from models of two schemes and CBH from CloudSat on CloudSat track
Channel number | Center wavelength /μm | Spatial resolution /km |
---|
1 | 0.47 | 1.0 | 2 | 0.65 | 0.5 | 3 | 0.825 | 1.0 | 4 | 1.375 | 2.0 | 5 | 1.61 | 2.0 | 6 | 2.25 | 2.0 | 7 | 3.75(H) | 2.0 | 8 | 3.75(L) | 4.0 | 9 | 6.25 | 4.0 | 10 | 7.1 | 4.0 | 11 | 8.5 | 4.0 | 12 | 10.7 | 4.0 | 13 | 12.0 | 4.0 | 14 | 13.5 | 4.0 |
|
Table 1. Parameters of FY-4A/AGRI channels
Cloud product of FY-4A/AGRI | Center wavelength of optical channel /μm |
---|
Cloud top height(CTH) | 11.0,12.09,13.55 | Cloud optical thickness(COT)and cloud effective radius(CER) | 0.65,2.25 |
|
Table 2. Related cloud products of FY-4A/AGRI and corresponding channels
Cloud type | Number |
---|
Ci | 73255 | As | 71000 | Ac | 41412 | St/Sc | 104792 | Cu | 24535 | Ns | 40397 | Dc | 10719 | Multi | 65603 |
|
Table 3. Numbers of all types of clouds after data matching
Cloud type | Decision trees number of RF model | Decision trees number of GBT model |
---|
Ci | 13 | 93 | As | 9 | 71 | Ac | 13 | 83 | St/Sc | 11 | 7 | Cu | 13 | 7 | Ns | 13 | 7 | Dc | 5 | 3 | Multi | 11 | 27 |
|
Table 4. Decision trees number of two models for all types of clouds
Cloud type | Maximum depth in RF | Maximum depth in GBT |
---|
Ci | 8 | 5 | As | 8 | 5 | Ac | 7 | 4 | St/Sc | 9 | 6 | Cu | 4 | 5 | Ns | 8 | 8 | Dc | 5 | 6 | Multi | 8 | 5 |
|
Table 5. Maximum depth of decision trees of two models for all types of clouds
Cloud | Mean value /m | RF model | GBT model |
---|
RMSE /m | MAE /m | R | MRE /% | RMSE /m | MAE /m | R | MRE /% |
---|
Ci | 9238.5 | 1060.6 | 841.4 | 0.7249 | 9.47 | 1054.8 | 836.5 | 0.7285 | 9.41 | As | 4918.6 | 1524.5 | 1180.9 | 0.6481 | 41.82 | 1515.8 | 1173.7 | 0.6532 | 41.65 | Ac | 2954.9 | 1357.2 | 1063.7 | 0.6396 | 66.68 | 1352.2 | 1060.0 | 0.6431 | 66.51 | St/Sc | 973.2 | 620.0 | 372.5 | 0.6062 | 43.45 | 669.5 | 394.4 | 0.5921 | 46.06 | Cu | 1075.6 | 1000.3 | 634.4 | 0.5018 | 78.63 | 1035.0 | 663.5 | 0.5132 | 84.22 | Ns | 1266.4 | 1338.8 | 896.2 | 0.4246 | 106.35 | 1371.7 | 933.2 | 0.4210 | 113.18 | Dc | 712.9 | 534.5 | 369.5 | 0.1847 | 60.12 | 538.4 | 375.6 | 0.1831 | 61.73 | Multi | 2285.0 | 2044.7 | 1593.7 | 0.4046 | 148.92 | 2044.9 | 1605.5 | 0.4069 | 150.72 |
|
Table 6. Retrieval results of two models for all types of clouds on the test dataset
Cloud type | Retrieval model |
---|
Ci | GBT | As | GBT | Ac | GBT | St/Sc | RF | Cu | RF | Ns | RF | Dc | RF | Multi | RF |
|
Table 7. Optimal CBH retrieval model of all types of clouds
Model | RMSE /m | MAE /m | R | MRE /% |
---|
RF | 2113.4 | 1497.6 | 0.7769 | 124.60 | GBT | 2109.1 | 1498.6 | 0.7779 | 124.81 |
|
Table 8. Retrieval results of RF model and GBT model on test dataset
Cloud type | Mean value /m | RF model | GBT model |
---|
RMSE /m | MAE /m | R | MRE /% | RMSE /m | MAE /m | R | MRE /% |
---|
Ci | 9238.5 | 2528.7 | 2178.0 | 0.6489 | 23.80 | 2529.7 | 2181.5 | 0.6518 | 23.84 | As | 4918.6 | 1896.3 | 1532.0 | 0.6111 | 41.19 | 1891.8 | 1529.5 | 0.6140 | 41.04 | Ac | 2954.9 | 1637.8 | 1246.1 | 0.5566 | 63.07 | 1636.4 | 1247.3 | 0.5566 | 63.17 | St/Sc | 973.2 | 1098.2 | 536.6 | 0.2782 | 66.40 | 1095.6 | 537.7 | 0.2763 | 66.69 | Cu | 1075.6 | 1734.9 | 1080.0 | 0.2450 | 171.41 | 1725.9 | 1076.7 | 0.2458 | 170.77 | Ns | 1266.4 | 1980.8 | 1518.9 | 0.2353 | 246.69 | 1977.5 | 1527.6 | 0.2320 | 248.25 | Dc | 712.9 | 3090.6 | 2524.6 | -0.0628 | 545.12 | 3071.9 | 2532.3 | -0.0578 | 546.44 | Multi | 2285.0 | 3104.5 | 2369.9 | 0.3775 | 297.83 | 3095.7 | 2367.4 | 0.3783 | 297.88 |
|
Table 9. Retrieval results of RF model and GBT model for all types of clouds on test dataset
Cloud | Mean value /m | Scheme one | Scheme two |
---|
RMSE /m | MAE /m | R | MRE /% | RMSE /m | MAE /m | R | MRE /% |
---|
Ci | 9238.5 | 1054.8 | 836.5 | 0.7285 | 9.41 | 2529.7 | 2181.5 | 0.6518 | 23.84 | As | 4918.6 | 1515.8 | 1173.7 | 0.6532 | 41.65 | 1891.8 | 1529.5 | 0.6140 | 41.04 | Ac | 2954.9 | 1352.2 | 1060.0 | 0.6431 | 66.51 | 1636.4 | 1247.3 | 0.5566 | 63.17 | St/Sc | 973.2 | 620.0 | 372.5 | 0.6062 | 43.45 | 1095.6 | 537.7 | 0.2763 | 66.69 | Cu | 1075.6 | 1000.3 | 634.4 | 0.5018 | 78.63 | 1725.9 | 1076.7 | 0.2458 | 170.77 | Ns | 1266.4 | 1338.8 | 896.2 | 0.4246 | 106.35 | 1977.5 | 1527.6 | 0.2320 | 248.25 | Dc | 712.9 | 534.5 | 369.5 | 0.1847 | 60.12 | 3071.9 | 2532.3 | -0.0578 | 546.44 | Multi | 2285.0 | 2044.7 | 1593.7 | 0.4046 | 148.92 | 3095.7 | 2367.4 | 0.3783 | 297.88 |
|
Table 10. Retrieval results of models of two schemes for all types of clouds on test dataset