Correlation Analysis between Air Temperature and MODIS Land Surface Temperature and Prediction of Air Temperature Using TensorFlow Long Short-Term Memory for the Period of Occurrence of Cold and Heat Waves

Lee, Yonggwan; Jang, Won-Jin; Lee, Siwoon; Kim, Seong-Joon

doi:10.3390/rs12193231

Cited by 24 publications

(12 citation statements)

References 46 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Although many studies use microwave radiometers to monitor regional or global SMC [12][13][14], there is a limit to detailed SMC or drought monitoring in a local area because of the coarse spatial resolution (~10-40 km) of microwave radiometers [15,16]. In the case of South Korea, 64% of the land area is made up of forests, and there is great topographical variability [17], so SMC monitoring using microwave radiometers with a large spatial resolution is not suitable. Active microwave synthetic aperture radars (SARs), such as ERS-1/2, RADARSAT-1, and JERS-1, have high resolutions of up to a few meters, but they are greatly affected by the surface roughness and vegetation cover, and the revisit cycle is generally longer than 15 days [18].…”

Section: Introductionmentioning

confidence: 99%

Soil Moisture Content Estimation Based on Sentinel-1 SAR Imagery Using an Artificial Neural Network and Hydrological Components

Lee

Kim

Jung³

et al. 2022

Remote Sensing

Self Cite

View full text Add to dashboard Cite

this study estimates soil moisture content (SMC) using Sentinel-1A/B C-band synthetic aperture radar (SAR) images and an artificial neural network (ANN) over a 40 × 50-km2 area located in the Geum River basin in South Korea. The hydrological components characterized by the antecedent precipitation index (API) and dry days were used as input data as well as SAR (cross-polarization (VH) and copolarization (VV) backscattering coefficients and local incidence angle), topographic (elevation and slope), and soil (percentage of clay and sand)-related data in the ANN simulations. A simple logarithmic transformation was useful in establishing the linear relationship between the observed SMC and the API. In the dry period without rainfall, API did not decrease below 0, thus the Dry days were applied to express the decreasing SMC. The optimal ANN architecture was constructed in terms of the number of hidden layers, hidden neurons, and activation function. The comparison of the estimated SMC with the observed SMC showed that the Pearson’s correlation coefficient (R) and the root mean square error (RMSE) were 0.85 and 4.59%, respectively.

show abstract

Section: Introductionmentioning

confidence: 99%

Soil Moisture Content Estimation Based on Sentinel-1 SAR Imagery Using an Artificial Neural Network and Hydrological Components

Lee

Kim

Jung³

et al. 2022

Remote Sensing

Self Cite

View full text Add to dashboard Cite

show abstract

“…The supply capacity ratio of agricultural water was 53.2-84.5% for the 41-year period. The agricultural water shortage exceeded 40% in 1988, 1994, 1995, and 2015, and South Korea was reported to have experienced a severe drought [39][40][41]. In 1988, the annual average precipitation and total runoff were 898.3 mm and 396.5 mm, respectively, and the agricultural water shortage was calculated as 512.5 × 10 6 m 3 .…”

Section: Actual Water Supply Analysis Results Of Modsim-dssmentioning

confidence: 99%

Evaluation of Agricultural Water Supply and Selection of Deficient Districts in Yeongsan River Basin of South Korea Considering Supply Priority

Kim

et al. 2022

Water

Self Cite

View full text Add to dashboard Cite

Considering the priority of water supply, agricultural water supply capacity downstream of the Yeongsan River Basin was evaluated in this study. The water balance was analyzed using the Please check all author names carefullyMODSIM-decision support system (MODSIM-DSS) to evaluate the agricultural water supply capacity. MODSIM-DSS can also be used to analyze the watershed demand status and agricultural water supply facilities. In this study, the watershed inflow and agricultural reservoir inflow (supply) data for each watershed were obtained using the SWAT outflow data. SWAT was calibrated and validated using 16-years (2005–2020) of daily streamflow data from one water level station and two weirs, by considering water withdrawal and return flows from agricultural, domestic, and industrial water use. The coefficient of determination (R2) was analyzed as 0.50 to 0.80 at three stations. Then, water balance analyses were performed for 41 years (1980–2020) by applying the SWAT outflow results to MODSIM-DSS. The results showed that agricultural water shortages of 517.8 × 106 m3, 520.0 × 106 m3, 579.8 × 106 m3, and 517.5 × 106 m3, occurred in 1992, 1994, 1995, and 2015, respectively. The efficiencies of agricultural water supply for these years were 58.2%, 58.0%, 53.2%, and 58.2%, respectively, which were lower than the 40-year average of 84.5%.

show abstract

“…Methods to obtain LST from spaceborne radiometers account for various effects related to surface properties and radiative characteristics of the atmosphere, but generally, LST values are found in good agreement with in situ measurements of soil temperatures (Gallo et al 2011;Good et al 2017). Nevertheless, the relationship between LST and air temperature is not straightforward: at large-scale and medium resolution, good correlations are found in the night with minimum air temperature, at a minor level also during the day with air temperature maxima (Chung et al 2020), while some authors do not consider viable to obtain air temperatures from satellite data (Xiong and Chen 2017). Moreover, in urban areas, the relationship is further complicated by enhanced surface heterogeneity, which gives rise to relevant and variable horizontal gradients in surface properties and affects the offset between in situ air temperature measurements and satellitederived LST (Elmes et al 2020;Sun et al 2020).…”

Section: Methodsmentioning

confidence: 95%

High-resolution climatic characterization of air temperature in the urban canopy layer

Montoli¹,

Frustaci²,

Lavecchia³

et al. 2021

Bull. of Atmos. Sci.& Technol.

View full text Add to dashboard Cite

Urbanized environments are of greater relevance because of the high and still rapidly increasing percentage of the world population living in and around cities and as the preferred location of human activities of every type. For this reason, much attention is paid to the urban climate worldwide. Among the UN 2030 17 Sustainable Development Goals, at least one concerns resilient cities and climate action. The WMO supports these goals promoting safe, healthy, and resilient cities by developing specially tailored integrated urban weather, climate, and environmental services. An unavoidable basis for that is an improved observational capability of urban weather and climate, as well as high-resolution modeling. For both the former and the latter, and of primary importance for the latter, urban meteorological surface networks are undoubtedly a very useful basis. Nevertheless, they are often unfit for detailed urban climatological studies and they are generally unable to describe the air temperature field in the urban canopy layer (UCL) with a spatial resolution which is sufficient to satisfy the requirements set by several professional activities and especially for local adaptation measures to climate change. On the other hand, remote sensing data from space offer a much higher spatial resolution of the surface characteristics, although the frequency is still relatively lower. A useful climatological variable from space is, for instance, the land surface temperature (LST), one of the WMO Essential Climate Variables (ECV). So often used to describe the Surface Urban Heat Islands (S-UHI), LST has no simple correlation with UCL air temperature, which is the most crucial variable for planning and management purposes in cities. In this work, after a review of correlation and interpolation methods and some experimentation, the cokriging methodology to obtain surface air temperature is proposed. The implemented methodology uses high quality but under-sampled in situ measurements of air temperature at the top of UCL, obtained by using a dedicated urban network, and satellite-derived LST. The satellite data used are taken at medium (1 × 1 km 2 ) resolution from Copernicus Sentinel 3 and at high resolution (30 × 30 m 2 ) from NASA-USGS Landsat 8. This fully exportable cokriging-based methodology, which also provides a quantitative measure of the related uncertainties, was tested and used to obtain medium to high spatial resolution air temperature maps of Milan (Italy) and the larger, much populated, but also partly rural, surrounding area of about 6000 km 2 . Instantaneous as well as long period mean fields of fine spatially resolved air temperature obtained by this method for selected weather types and different Urban Heat Island configurations represent an important knowledge improvement for the climatology of the urban Extended author information available on the last page of the article Bulletin of Atmospheric Science and Technology 1 3 and peri-urban area of Milan. It finds application not only in more detailed urban climat...

show abstract

Correlation Analysis between Air Temperature and MODIS Land Surface Temperature and Prediction of Air Temperature Using TensorFlow Long Short-Term Memory for the Period of Occurrence of Cold and Heat Waves

Cited by 24 publications

References 46 publications

Soil Moisture Content Estimation Based on Sentinel-1 SAR Imagery Using an Artificial Neural Network and Hydrological Components

Soil Moisture Content Estimation Based on Sentinel-1 SAR Imagery Using an Artificial Neural Network and Hydrological Components

Evaluation of Agricultural Water Supply and Selection of Deficient Districts in Yeongsan River Basin of South Korea Considering Supply Priority

High-resolution climatic characterization of air temperature in the urban canopy layer

Contact Info

Product

Resources

About