Localization or Globalization? Determination of the Optimal Regression Window for Disaggregation of Land Surface Temperature

Gao, Lun; Zhan, Wenfeng; Huang, Fan; Quan, Jinling; Lu, Xiaoman; Wang, Fei; Ju, Weimin; Zhou, Ji

doi:10.1109/tgrs.2016.2608987

Cited by 20 publications

(29 citation statements)

References 44 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…When the window size is too large, the heterogeneities among the pixels in the window increase, making the relationship between the kernels and LST more complicated and thus decreasing the accuracy of DLST. Moreover, a larger window that includes more training samples increases the computational complexity [30]. Hence, an optimal segmentation scale that considers the accuracy and computational cost is required for the OWS.…”

Section: Object-based Window Strategymentioning

confidence: 99%

“…However, the optimal window size for an arbitrary area was not determined since a large window might contain more heterogeneous pixels that reduce the stability of the regression relationship, while a small window might be time-consuming due to the large number of regressions. Hence, Gao [30] proposed an indirect criterion based on aggregation-disaggregation (ICAD), with which we can choose between the global window strategy (GWS) and the local window strategy (LWS) and determine the optimal moving window size of the LWS considering both accuracy and efficiency.…”

Section: Introductionmentioning

confidence: 99%

“…The moving window used in the LWS is square, which may be less sufficient than a circular window since a circular window defines better correlations among nearby pixels [30]. Furthermore, the shape of land cover objects is not always regular, and even a circular window cannot always derive stable correlations.…”

Section: Introductionmentioning

confidence: 99%

“…Furthermore, the shape of land cover objects is not always regular, and even a circular window cannot always derive stable correlations. Considering local anisotropic heterogeneity, the best regression window should be determined by homogeneous pixels, which can be any shape beyond a square or a circle [30]. Object-based windows derived from segmentation algorithms may generate more stable correlations between the LST and kernels considering that the pixels within the window are homogeneous.…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Object-Based Window Strategy in Thermal Sharpening

et al. 2019

Self Cite

View full text Add to dashboard Cite

The trade-off between spatial and temporal resolutions has led to the disaggregation of remotely sensed land surface temperatures (LSTs) for better applications. The window used for regression is one of the primary factors affecting the disaggregation accuracy. Global window strategies (GWSs) and local window strategies (LWSs) have been widely used and discussed, while object-based window strategies (OWSs) have rarely been considered. Therefore, this study presents an OWS based on a segmentation algorithm and provides a basis for selecting an optimal window size balancing both accuracy and efficiency. The OWS is tested with Landsat 8 data and simulated data via the “aggregation-then-disaggregation” strategy, and compared with the GWS and LWS. Results tested with the Landsat 8 data indicate that the proposed OWS can accurately and efficiently generate high-resolution LSTs. In comparison to the GWS, the OWS improves the mean accuracy by 0.19 K at different downscaling ratios, in particular by 0.30 K over urban areas; compared with the LWS, the OWS performs better in most cases but performs slightly worse due to the increasing downscaling ratio in some cases. Results tested with the simulated data indicate that the OWS is always superior to both GWS and LWS regardless of the downscaling ratios, and the OWS improves the mean accuracy by 0.44 K and 0.19 K in comparison to the GWS and LWS, respectively. These findings suggest the potential ability of the OWS to generate super-high-resolution LSTs over heterogeneous regions when the pixels within the object-based windows derived via segmentation algorithms are more homogenous.

show abstract

Section: Object-based Window Strategymentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Object-Based Window Strategy in Thermal Sharpening

et al. 2019

Self Cite

View full text Add to dashboard Cite

show abstract

“…Generally, most have been focused on the relationship between low-resolution TIR and high-resolution vegetation index (VI), especially normalized difference vegetation index (NDVI), because densely vegetated areas tend to have relatively low surface temperature [6][7][8][9][10][11][12][13][14][15]. Disaggregation procedures for radiometric surface temperature (DisTrad) [6] and thermal sharpening (TsHARP) [7] algorithms are viewed as the leading methods due to their effectiveness and simplicity.…”

Section: Introductionmentioning

confidence: 99%

Disaggregation of Landsat-8 Thermal Data Using Guided SWIR Imagery on the Scene of a Wildfire

Cho

Kim

2018

Remote Sensing

View full text Add to dashboard Cite

Thermal data products derived from remotely sensed data play significant roles as key parameters for biophysical phenomena. However, a trade-off between spatial and spectral resolutions has existed in thermal infrared (TIR) remote sensing systems, with the end product being the limited resolution of the TIR sensor. In order to treat this problem, various disaggregation methods of TIR data, based on the indices from visible and near-infrared (VNIR), have been developed to sharpen the coarser spatial resolution of TIR data. Although these methods were reported to exhibit sufficient performance in each study, preservation of thermal variation in the original TIR data is still difficult, especially in fire areas due to the distortion of the VNIR reflectance by the impact of smoke. To solve this issue, this study proposes an efficient and improved disaggregation algorithm of TIR imagery on wildfire areas using guided shortwave infrared (SWIR) band imagery via a guided image filter (GF). Radiometric characteristics of SWIR wavelengths could preserve spatially high frequency temperature components in flaming combustion, and the GF preserved thermal variation of the original TIR data in the disaggregated result. The proposed algorithm was evaluated using Landsat-8 operational land imager (OLI) and thermal infrared sensor (TIRS) images on wildfire areas, and compared with other algorithms based on a vegetation index (VI) originating from VNIR. In quantitative analysis, the proposed disaggregation method yielded the best values of root mean square error (RMSE), mean absolute error (MAE), correlation coefficient (CC), erreur relative globale adimensionelle de synthèse (ERGAS), and universal image quality index (UIQI). Furthermore, unlike in other methods, the disaggregated temperature map in the proposed method reflected the thermal variation of wildfire in visual analysis. The experimental results showed that the proposed algorithm was successfully applied to the TIR data, especially to wildfire areas in terms of quantitative and visual assessments.

show abstract