Mapping and Monitoring Rice Agriculture with Multisensor Temporal Mixture Models

Abstract

Rice is the staple food for more than half of humanity. Accurate prediction of rice harvests is therefore of considerable global importance for food security and economic stability, especially in the developing world. Landsat sensors have collected coincident thermal and optical images for the past 35+ years, and so can provide both retrospective and near-realtime constraints on the spatial extent of rice planting and the timing of rice phenology. Thermal and optical imaging capture different physical processes, and so provide different types of information for phenologic mapping. Most analyses use only one or the other data source, omitting potentially useful information. We present a novel approach to the mapping and monitoring of rice agriculture which leverages both optical and thermal measurements. The approach relies on Temporal Mixture Models (TMMs) derived from parallel Empirical Orthogonal Function (EOF) analyses of Landsat image time series. Analysis of each image time series is performed in two stages: (1) spatiotemporal characterization, and (2) temporal mixture modeling. Characterization evaluates the covariance structure of the data, culminating in the selection of temporal endmembers (EMs) representing the most distinct phenological cycles of either vegetation abundance or surface temperature. Modeling uses these EMs as the basis for linear TMMs which map the spatial distribution of each EM phenological pattern across study area. The two metrics we analyze in parallel are (1) fractional vegetation abundance (Fv) derived from spectral mixture analysis (SMA) of optical reflectance, and (2) land surface temperature (LST) derived from brightness temperature (Tb). These metrics are chosen on the basis of being straightforward to compute for any (cloud-free) Landsat 4-8 image in the global archive. We demonstrate the method using a 90 × 120 km area in the Sacramento Valley of California. Satellite Tb retrievals are corrected to LST using a standardized atmospheric correction approach and pixelwise fractional emissivity estimates derived from SMA. LST and Tb time series are compared to field station data in 2016 and 2017. Uncorrected Tb is observed to agree with the upper bound of the envelope of air temperature observations to within 3 °C on average. As expected, LST estimates are 3 to 5 °C higher. Soil T, air T, Tb and LST estimates can all be represented as linear transformations of the same seasonal cycle. The 3D temporal feature spaces of Fv and LST clearly resolve 5 and 7 temporal EM phenologies, respectively, with strong clustering distinguishing rice from other vegetation. Results from parallel EOF analyses of coincident Fv and LST image time series over the 2016 and 2017 growing seasons suggest that TMMs based on single year Fv datasets can provide accurate maps of crop timing, while TMMs based on dual year LST datasets can provide comparable maps of year-to-year crop conversion. We also test a partial-year model midway through the 2018 growing season to illustrate a potential real-time monitoring application. Field validation confirms the monitoring model provides an upper bound estimate of spatial extent and relative timing of the rice crop accurate to 89%, even with an unusually sparse set of usable Landsat images.

Land Surface Temperature Retrieval for Agricultural Areas Using a Novel UAV Platform Equipped with a Thermal Infrared and Multispectral Sensor

“…Different assumptions were made to emphasise the effects of when essential parameters, such as transmittance and environmental thermal emissions, were not taken into account in LST retrieval (Section 2.1.3). Each case was compared to the uncorrected BT sens (Figure 7a) and the modified LST retrievals (Equations (11) and (12)) to depict the consequences of inaccurately retrieved LSTs: Case i: Atmospheric transmittance included in LST(τ), but without consideration of the background temperature and specific emissivity (Equation (11)); and case ii: The background temperature and specific emissivity included in LST( ), but without consideration of transmittance (Equation (12)). For BT sens without consideration of the atmospheric transmittance, a temperature deviance of up to 0.4 K occurred (Figure 8a).…”

“…Based on these limiting assumptions (Equations (11) and (12)), we made the following two case distinctions: Case i: LST(τ) includes the atmospheric transmittance, but does not take into account the emissivity or background temperature (Equation (11)); and case ii: LST( ) includes the emissivity and background temperature, but omits the transmittance (Equation 12). In summary, we have highlighted the potential implications of omitting these parameters, which should be considered in order to minimise the uncertainties in retrieving precise LST values.…”

“…Therefore, the LST is a key parameter for the physical description of the surface energy and water balance processes at the local to global scale [1,2]. The potential of this crucial parameter has been repeatedly demonstrated in various thermal infrared-based studies and applications, such as evapotranspiration [3,4], hydrological modelling [5], vegetation monitoring [6], 'urban heat island and urban development' [7][8][9], climate change and weather conditions [1,10], agriculture [3,11], and the monitoring of land use changes in wetlands [12]. In agricultural applications, precision farming has increasingly required thermal remote sensing techniques to detect water-stressed crops [3,13,14], plant diseases [13,15], and for irrigation management [13,14].…”

Year-Independent Prediction of Food Insecurity Using Classical and Neural Network Machine Learning Methods

“…Current prediction efforts rely on a variety of demographic, geographic, and societal attributes but are not able to anticipate real-time shortfalls in crop production and landscape changes. To answer this, machine learning algorithms have recently been employed to predict crop yields, and remote sensing techniques are being evaluated to monitor drought and salinity stress simultaneously [7][8][9]. Thus, governments must continue to increase the fidelity with which they measure and respond to food insecurity.…”

Remote sensing versus the area sampling frame method in paddy rice acreage estimation in Indramayu regency, West Java province, Indonesia

“…Another means for producing rice data is through the use of phenology-based remote sensing with the ability to map and monitor rice fields on a wide scale quickly and accurately. This method has been applied with accuracy of 82% to 95% in various countries including in the United States (Sousa and Small 2019), Turkey (Rufin et al 2019), France (Bazzi et al 2019), China (Mansaray et al 2019), India, Vietnam, Cambodia, Thailand, the Philippines and Indonesia (Nelson et al 2014). In tropical island countries such as Indonesia, with relatively high air humidity (an average of more than 90%), high rainfall, high temperatures and heavy cloud, medium and long wave remote sensing SAR methods are believed to be more effective than ground-based approaches (Mosleh, Hassan, and Chowdhury 2015;Mansaray et al 2019).…”