Classification and Mapping of Paddy Rice by Combining Landsat and SAR Time Series Data

Park, Seonyoung; Im, Jungho; Park, Seohui; Yoo, Cheolhee; Han, Hyangsun; Rhee, Jinyoung

doi:10.3390/rs10030447

Cited by 116 publications

(68 citation statements)

References 77 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…In fact, the choice of a regressor among various machine learning options significantly affects the prediction results (Lee et al, 2018; Liu et al, 2018; Park et al, 2018; Wylie et al, 2019). Although several machine learning algorithms have already been used in temperature bias‐correction, improving the modeling accuracy remains challenging.…”

Section: Introductionmentioning

confidence: 99%

Comparative Assessment of Various Machine Learning‐Based Bias Correction Methods for Numerical Weather Prediction Model Forecasts of Extreme Air Temperatures in Urban Areas

Cho

Yoo

et al. 2020

Earth and Space Science

Self Cite

110

View full text Add to dashboard Cite

Forecasts of maximum and minimum air temperatures are essential to mitigate the damage of extreme weather events such as heat waves and tropical nights. The Numerical Weather Prediction (NWP) model has been widely used for forecasting air temperature, but generally it has a systematic bias due to its coarse grid resolution and lack of parametrizations. This study used random forest (RF), support vector regression (SVR), artificial neural network (ANN) and a multi-model ensemble (MME) to correct the Local Data Assimilation and Prediction System (LDAPS; a local NWP model over Korea) model outputs of next-day maximum and minimum air temperatures (T maxtþ1 and T mintþ1 ) in Seoul, South Korea. A total of 14 LDAPS model forecast data, the daily maximum and minimum air temperatures of in-situ observations, and five auxiliary data were used as input variables. The results showed that the LDAPS model had an R 2 of 0.69, a bias of −0.85°C and an RMSE of 2.08°C for T maxtþ1 forecast, whereas the proposed models resulted in the improvement with R 2 from 0.75 to 0.78, bias from −0.16 to −0.07°C and RMSE from 1.55 to 1.66°C by hindcast validation. For forecasting T mintþ1 , the LDAPS model had an R 2 of 0.77, a bias of 0.51°C and an RMSE of 1.43°C by hindcast, while the bias correction models showed R 2 values ranging from 0.86 to 0.87, biases from −0.03 to 0.03°C, and RMSEs from 0.98 to 1.02°C. The MME model had better generalization performance than the three single machine learning models by hindcast validation and leave-one-station-out cross-validation.

show abstract

Section: Introductionmentioning

confidence: 99%

Comparative Assessment of Various Machine Learning‐Based Bias Correction Methods for Numerical Weather Prediction Model Forecasts of Extreme Air Temperatures in Urban Areas

Cho

Yoo

et al. 2020

Earth and Space Science

Self Cite

110

View full text Add to dashboard Cite

show abstract

“…The Moderate Resolution Imaging Spectroradiometer (MODIS) sensors have been one of the most widely used platforms for rice mapping and monitoring applications at larger regional scales because of the daily revisit, relatively small data size, and availability of spectral information that is particularly pertinent to agriculture [9,10,23,26,27]. MODIS time series images have also been combined with synthetic aperture radar (SAR) data for rice monitoring [28,29], exemplifying new initiatives and innovative techniques that are becoming available with the advent of free access to remotely sensed datasets and improved information on regional rice production systems. However, MODIS has only been available since the year 2000, and its moderate spatial resolution is not suitable for heterogeneous landscapes such as Bangladesh, where the average farm size is less than one quarter of a hectare [30].…”

Section: Introductionmentioning

confidence: 99%

Identifying Dry-Season Rice-Planting Patterns in Bangladesh Using the Landsat Archive

Shew

Ghosh

2019

Remote Sensing

View full text Add to dashboard Cite

In many countries, in situ agricultural data is not available and cost-prohibitive to obtain. While remote sensing provides a unique opportunity to map agricultural areas and management characteristics, major efforts are needed to expand our understanding of cropping patterns and the potential for remotely monitoring crop production because this could support predictions of food shortages and improve resource allocation. In this study, we demonstrate a new method to map paddy rice using Google Earth Engine (GEE) and the Landsat archive in Bangladesh during the dry (boro) season. Using GEE and Landsat, dry-season rice areas were mapped at 30 m resolution for approximately 90,000 km2 annually between 2014 and 2018. The method first reconstructs spectral vegetation indices (VIs) for individual pixels using a harmonic time series (HTS) model to minimize the effect of any sensor inconsistencies and atmospheric noise, and then combines the time series indices with a rule-based algorithm to identify characteristics of rice phenology to classify rice pixels. To our knowledge, this is the first time an annual pixel-based time series model has been applied to Landsat at the national level in a multiyear analysis of rice. Findings suggest that the harmonic-time-series-based vegetation indices (HTS-VIs) model has the potential to map rice production across fragmented landscapes and heterogeneous production practices with comparable results to other estimates, but without local management or in situ information as inputs. The HTS-VIs model identified 4.285, 4.425, 4.645, 4.117, and 4.407 million rice-producing hectares for 2014, 2015, 2016, 2017, and 2018, respectively, which correlates well with national and district estimates from official sources at an average R-squared of 0.8. Moreover, accuracy assessment with independent validation locations resulted in an overall accuracy of 91% and a kappa coefficient of 0.83 for the boro/non-boro stable rice map from 2014 to 2018. We conclude with a discussion of potential improvements and future research pathways for this approach to spatiotemporal mapping of rice in heterogeneous landscapes.

show abstract

“…All these optical data are susceptible to weather conditions and image coverage, but reducing feasibility in rice-monitoring applications. Multiseasonal synthetic aperture radar (SAR) data, being immune to weather conditions, have better performance for rice monitoring [30][31][32][33]. The cost of SAR data rises rapidly with the increase of resolution [17,34].…”

mentioning

confidence: 99%

Mapping Paddy Rice Using a Convolutional Neural Network (CNN) with Landsat 8 Datasets in the Dongting Lake Area, China

et al. 2018

View full text Add to dashboard Cite

Rice is one of the world’s major staple foods, especially in China. Highly accurate monitoring on rice-producing land is, therefore, crucial for assessing food supplies and productivity. Recently, the deep-learning convolutional neural network (CNN) has achieved considerable success in remote-sensing data analysis. A CNN-based paddy-rice mapping method using the multitemporal Landsat 8, phenology data, and land-surface temperature (LST) was developed during this study. First, the spatial–temporal adaptive reflectance fusion model (STARFM) was used to blend the moderate-resolution imaging spectroradiometer (MODIS) and Landsat data for obtaining multitemporal Landsat-like data. Subsequently, the threshold method is applied to derive the phenological variables from the Landsat-like (Normalized difference vegetation index) NDVI time series. Then, a generalized single-channel algorithm was employed to derive LST from the Landsat 8. Finally, multitemporal Landsat 8 spectral images, combined with phenology and LST data, were employed to extract paddy-rice information using a patch-based deep-learning CNN algorithm. The results show that the proposed method achieved an overall accuracy of 97.06% and a Kappa coefficient of 0.91, which are 6.43% and 0.07 higher than that of the support vector machine method, and 7.68% and 0.09 higher than that of the random forest method, respectively. Moreover, the Landsat-derived rice area is strongly correlated (R2 = 0.9945) with government statistical data, demonstrating that the proposed method has potential in large-scale paddy-rice mapping using moderate spatial resolution images.

show abstract

Classification and Mapping of Paddy Rice by Combining Landsat and SAR Time Series Data

Cited by 116 publications

References 77 publications

Comparative Assessment of Various Machine Learning‐Based Bias Correction Methods for Numerical Weather Prediction Model Forecasts of Extreme Air Temperatures in Urban Areas

Comparative Assessment of Various Machine Learning‐Based Bias Correction Methods for Numerical Weather Prediction Model Forecasts of Extreme Air Temperatures in Urban Areas

Identifying Dry-Season Rice-Planting Patterns in Bangladesh Using the Landsat Archive

Mapping Paddy Rice Using a Convolutional Neural Network (CNN) with Landsat 8 Datasets in the Dongting Lake Area, China

Contact Info

Product

Resources

About