Diurnal Air Temperature Modeling Based on the Land Surface Temperature

Mehdi Gholamnia, Ali Darvishi Boloorani, Saeid Hamzeh, Majid Kiavarz
2017 Remote Sensing  
The air temperature is an essential variable in many applications related to Earth science. Sporadic spatial distribution of weather stations causes a low spatial resolution of measured air temperatures. This study focused on modeling the air diurnal temperature cycle (DTC) based on the land surface temperature (LST) DTC. The air DTC model parameters were estimated from LST DTC model parameters by a regression analysis. Here, the LST obtained from the INSAT-3D geostationary satellite and the
more » ... temperature extracted from weather stations were used within the time frame of 4 March 2015 to 22 May 2017 across Iran. Constant parameters of the air DTC model for each weather station were estimated based on an experimental approach over the time period. Results showed these parameters decrease as elevation increases. The mean absolute error (MAE) and the root mean square error (RMSE) for three hours sampling were calculated. The MAE and RMSE ranges were between [0.1, 4] • C and [0.1, 3.3] • C, respectively. Additionally, 95% of MAEs and RMSEs were less than 2.9 • C and 2.4 • C values, correspondingly. The range of the mean values of MAEs and RMSEs for a three-hour sampling time were [−0.29, 0.6] • C and [2, 2.11] • C. The DTC model results showed a meaningful statistical fitting in both air DTCs modeled from LST and weather station-based DTCs. The variability of mean error and RMSE in different land covers and elevation classes were also investigated. In spite of the complex behavior of the environmental variables in the study area, the model error bar did not show significantly biased estimations for various classes. Therefore, the developed model was less sensitive to variations of land covers and elevation changes. It can be conclude that the coefficients of regression between LST and air DTC could model properly the environmental factors. Remote Sens. 2017, 9, 915 2 of 12 air temperature is complicated because it is influenced by a broad range of variable parameters, like wind speed, time, sky condition, solar zenith angle, soil moisture, location, albedo, emissivity, and thermal inertia [7] . Some of these parameters cannot be measured by remote sensing. Researchers are trying to develop methodologies to include the effects of these parameters in their modeling. Most of the developed methods are time-and location-dependent and are not applicable in other cases. According to Zakšek and Schroedter-Homscheidt [8], the developed methods are classified into three major categories: (1) Temperature-Vegetation Index (TVX) [9] [10] [11] [12] , which is based on the assumption that temperature of the vegetation canopy is almost equal to the ambient air temperature since their heat capacity is similar [7, 9, 10, 13] . This method is sensitive to seasonality, soil moisture, and land cover type [14, 15] ; (2) The surface energy balance principle [6, 7, [16] [17] [18] . These approaches are based on net radiation, sensible heat flux, latent heat flux, and ground heat flux. Zhu et al. [19] mentioned that these models need many inputs which are not fully retrievable from remote sensing data; (3) Statistical approaches which recruit parametric and non-parametric machine learning approaches [14, 16, 18, [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] . Such methods use ancillary data, such as Julian date, latitude, longitude, terrain, and vegetation indices to improve their modeling. Xu et al. [31] stated that linear regression cannot predict the air temperature based on LST in all conditions. They specified that Random Forest (RF) showed better accuracy than the multiple linear regression model. Air temperature modeling based on TVX and statistical methods are conducted based on discrete records of air temperatures in weather stations and the time between these records are not included in calculations. Furthermore, the effects of noise in recorded data caused by atmosphere forcing, cloud contamination, and other environmental conditions generate errors into the air temperature estimation. Diurnal modeling of LST and air temperature [32] can be beneficial in the modeling of Ta. Therefore, a synoptic coverage of Ta is obtainable. Jin and Dickinson [33] estimated the diurnal cycle of LST through temporal interpolation of NOAA-AVHRR. Sun et al. [34] used the same method for modeling Geostationary Operational Environmental Satellite (GOES) measurements. Alavipanah et al. [35] studied the application of LST DTC models for knowing the LST discrepancies between MODIS and geostationary satellites. Inamdar et al. [36] proposed an approach which employs MODIS imagery for calibration of GOES. They used both datasets to build half-hourly LSTs with 1 km spatial resolution. Stisen et al. [10] used TVX-based DTC modeling to acquire diurnal variation of air temperature. This research attempted to introduce methods that extract air temperature based on LST data by direct approach and without auxiliary environmental information. This study had three main goals: (1) estimating the LST DTC parameters in weather stations; (2) regression-driving of air temperature DTCs based on estimated parameters of LST DTCs; (3) comparing DTC parameters of LST and Ta; and (4) evaluation of the modeled air DTC model by weather station observations.
doi:10.3390/rs9090915 fatcat:ii5zwlbfmjf5nfcjoxfvhhlgty