scholarly journals Bayesian Sigmoid-Type Time Series Forecasting with Missing Data for Greenhouse Crops

Sensors ◽  
2020 ◽  
Vol 20 (11) ◽  
pp. 3246 ◽  
Author(s):  
Alexander Kocian ◽  
Giulia Carmassi ◽  
Fatjon Cela ◽  
Luca Incrocci ◽  
Paolo Milazzo ◽  
...  
Keyword(s):  
Time Series ◽  
Environmental Control ◽  
Measurement Data ◽  
Environmental Parameters ◽  
Integrated Approach ◽  
Crop Growth ◽  
Parameter Estimates ◽  
Sigmoid Function ◽  
Area Index ◽  
Hidden States

This paper follows an integrated approach of Internet of Things based sensing and machine learning for crop growth prediction in agriculture. A Dynamic Bayesian Network (DBN) relates crop growth associated measurement data to environmental control data via hidden states. The measurement data, having (non-linear) sigmoid-type dynamics, are instances of the two classes observed and missing, respectively. Considering that the time series of the logistic sigmoid function is the solution to a reciprocal linear dynamic model, the exact expectation-maximization algorithm can be applied to infer the hidden states and to learn the parameters of the model. At iterative convergence, the parameter estimates are then used to derive a predictor of the measurement data several days ahead. To evaluate the performance of the proposed DBN, we followed three cultivation cycles of micro-tomatoes (MicroTom) in a mini-greenhouse. The environmental parameters were temperature, converted into Growing Degree Days (GDD), and the solar irradiance, both at a daily granularity. The measurement data were Leaf Area Index (LAI) and Evapotranspiration (ET). Although measurement data were only available scarcely, it turned out that high quality measurement data predictions were possible up to three weeks ahead.

scholarly journals Assessing the Impact of Satellite Revisit Rate on Estimation of Corn Phenological Transition Timing through Shape Model Fitting

2019 ◽  
Vol 11 (21) ◽  
pp. 2558 ◽  
Author(s):  
Emily Myers ◽  
John Kerekes ◽  
Craig Daughtry ◽  
Andrew Russ
Keyword(s):  
Time Series ◽  
Vegetation Index ◽  
Time Series Data ◽  
Model Fitting ◽  
Series Data ◽  
Sigmoid Function ◽  
Absolute Deviation ◽  
Shape Model ◽  
Area Index ◽  
The Impact

Agricultural monitoring is an important application of earth-observing satellite systems. In particular, image time-series data are often fit to functions called shape models that are used to derive phenological transition dates or predict yield. This paper aimed to investigate the impact of imaging frequency on model fitting and estimation of corn phenological transition timing. Images (PlanetScope 4-band surface reflectance) and in situ measurements (Soil Plant Analysis Development (SPAD) and leaf area index (LAI)) were collected over a corn field in the mid-Atlantic during the 2018 growing season. Correlation was performed between candidate vegetation indices and SPAD and LAI measurements. The Normalized Difference Vegetation Index (NDVI) was chosen for shape model fitting based on the ground truth correlation and initial fitting results. Plot-average NDVI time-series were cleaned and fit to an asymmetric double sigmoid function, from which the day of year (DOY) of six different function parameters were extracted. These points were related to ground-measured phenological stages. New time-series were then created by removing images from the original time-series, so that average temporal spacing between images ranged from 3 to 24 days. Fitting was performed on the resampled time-series, and phenological transition dates were recalculated. Average range of estimated dates increased by 1 day and average absolute deviation between dates estimated from original and resampled time-series data increased by 1/3 of a day for every day of increase in average revisit interval. In the context of this study, higher imaging frequency led to greater precision in estimates of shape model fitting parameters used to estimate corn phenological transition timing.

Environmental control of potential yield of sunflower in the subtropics

1997 ◽  
Vol 48 (2) ◽  
pp. 231 ◽  
Author(s):  
M. P. Bange ◽  
G. L. Hammer ◽  
K. G. Rickert
Keyword(s):  
Harvest Index ◽  
Environmental Control ◽  
Grain Filling ◽  
Crop Growth ◽  
Light Interception ◽  
Radiation Environment ◽  
Potential Yield ◽  
Available Nitrogen ◽  
Area Index ◽  
High Yields

A simple framework was used to analyse the determinants of potential yield of sunflower (Helianthus annuus L.) in a subtropical environment. The aim was to investigate the stability of the determinants crop duration, canopy light interception, radiation use efficiency (RUE), and harvest index (HI) at 2 sowing times and with 3 genotypes differing in crop maturity and stature. Crop growth, phenology, light interception, yield, prevailing temperature, and radiation were recorded and measured throughout the crop cycle. Significant differences in grain yield were found between the 2 sowings, but not among genotypes within each sowing. Mean yields (0% moisture) were 6·02 and 2·17 t/ha for the first sowing, on 13 September (S1), and the second sowing, on 5 March (S2), respectively. Exceptionally high yields in S1 were due to high biomass assimilation associated with the high radiation environment, high light interception owing to a greater leaf area index, and high RUE (1·47–1·62 g/MJ) across genotypes. It is proposed that the high RUE was caused by high levels of available nitrogen maintained during crop growth by frequent applications of fertiliser and sewage effluent as irrigation. In addition to differences in the radiation environment, the assimilate partitioned to grain was reduced in S2 associated with a reduction in the duration of grain-filling. Harvest index was 0·40 in S1 and 0·25 in S2. It is hypothesised that low minimum temperatures experienced in S2 reduced assimilate production and partitioning, causing premature maturation.

scholarly journals Integration of a Crop Growth Model and Deep Learning Methods to Improve Satellite-Based Yield Estimation of Winter Wheat in Henan Province, China

2021 ◽  
Vol 13 (21) ◽  
pp. 4372
Author(s):  
Yi Xie ◽  
Jianxi Huang
Keyword(s):  
Time Series ◽  
Deep Learning ◽  
Crop Yield ◽  
Growth Model ◽  
Wheat Yield ◽  
Crop Growth ◽  
Henan Province ◽  
Crop Growth Model ◽  
Area Index ◽  
Growing Region

Timely and accurate regional crop-yield estimates are crucial for guiding agronomic practices and policies to improve food security. In this study, a crop-growth model was integrated with time series of remotely sensed data through deep learning (DL) methods to improve the accuracy of regional wheat-yield estimations in Henan Province, China. Firstly, the time series of moderate-resolution imaging spectroradiometer (MODIS) normalized difference vegetation index (NDVI) were input into the long short-term memory network (LSTM) model to identify the wheat-growing region, which was further used to estimate wheat areas at the municipal and county levels. Then, the leaf area index (LAI) and grain-yield time series simulated by the Crop Environment REsource Synthesis for Wheat (CERES-Wheat) model were used to train and evaluate the LSTM, one-dimensional convolutional neural network (1-D CNN) and random forest (RF) models, respectively. Finally, an exponential model of the relationship between the field-measured LAI and MODIS NDVI was applied to obtain the regional LAI, which was input into the trained LSTM, 1-D CNN and RF models to estimate wheat yields within the wheat-growing region. The results showed that the linear correlations between the estimated wheat areas and the statistical areas were significant at both the municipal and county levels. The LSTM model provided more accurate estimates of wheat yields, with higher R2 values and lower root mean square error (RMSE) and mean relative error (MRE) values than the 1-D CNN and RF models. The LSTM model has an inherent advantage in capturing phenological information contained in the time series of the MODIS-derived LAI, which is important for satellite-based crop-yield estimates.

Monitoring crop growth and key agronomic parameters through multitemporal observations and time series analysis from remote sensing big data

2017 ◽  
Vol 8 (2) ◽  
pp. 394-399 ◽  
Author(s):  
K. Karantzalos ◽  
A. Karmas ◽  
A. Tzotsos
Keyword(s):  
Remote Sensing ◽  
Time Series ◽  
Big Data ◽  
Value Added ◽  
Remote Sensing Data ◽  
Growth Cycle ◽  
Crop Growth ◽  
Area Index ◽  
Growth State ◽  
Server Side

In this paper, novel geospatial services are presented which are able to process on the server-side numerous remote sensing data based on big data frameworks like Hadoop and Rasdaman. The developed system itself features several software modules that orchestrate the different image processing algorithms responsible for the production of consistent value-added maps like canopy greenness and leaf area index. Through distributed multitemporal analysis, the entire crop growth cycle can be continuously monitored through the analysis of time-series observations. These observations cover multiple crop growth cycles, offering invaluable information by linking weather statistical data with the start, the end and the duration of each growth cycle enabling critical decisions by direct comparison with the current crop growth state.

scholarly journals Combining hyperspectral remote sensing and eddy covariance data streams for estimation of vegetation functional traits

2020 ◽  
Author(s):  
Javier Pacheco-Labrador ◽  
Tarek S. El-Madany ◽  
M. Pilar Martin ◽  
Rosario Gonzalez-Cascon ◽  
Arnaud Carrara ◽  
...  
Keyword(s):  
Remote Sensing ◽  
Time Series ◽  
Eddy Covariance ◽  
Functional Traits ◽  
Hyperspectral Imagery ◽  
Observation System ◽  
Parameter Estimates ◽  
Area Index ◽  
Functional Parameters ◽  
New Generation

Abstract. Remote Sensing (RS) has traditionally provided estimates of key biophysical properties controlling light interaction with the canopy (e.g., chlorophyll content (Cab) or leaf area index (LAI)). However, recent and upcoming developments in hyperspectral RS are expected to lead to a new generation of products such as vegetation functional traits that control leaf carbon and water gas exchange. This information is pivotal to improve our understanding and capability to predict biosphere-atmosphere fluxes at global scale. Yet, the retrieval of key functional traits such as maximum carboxylation rate (Vcmax) or the Ball-Berry stomatal sensitivity parameter (m) remains challenging, as they only have a weak and indirect influence on optical reflectance factors. Recently, the assimilation of different observations in coupled soil-vegetation-atmosphere transfer (SVAT) and radiative transfer models (RTM) is allowing Vcmax and m estimates; notably using the Soil Canopy Observation of Photosynthesis and Energy fluxes (SCOPE) model. In this work we assess the potential of airborne and satellite emulated hyperspectral imagery jointly with eddy covariance (EC) data for the retrieval of functional traits. Specifically, we made use of time series of gross primary production (GPP) and thermal irradiance measured with net radiometers, together with 17 hyperspectral airborne images. The potential of satellite-borne sensors was tested with emulated EnMAP imagery from the airborne data. EnMAP was selected because of the availability of the emulator, and because is one of the foreseen hyperspectral satellite missions expected to contribute to a new generation of RS products. We estimated ecosystem functional traits by inverting the senSCOPE model, a novel version of SCOPE adapted to represent partly senescent canopies. The experiment takes place in a Mediterranean tree-grass ecosystem subject of a large scale manipulation experiment with nitrogen and nitrogen plus phosphorus, monitored by three EC towers. Parameter estimates and predicted fluxes were evaluated using both ground observations and pattern-oriented model evaluation approach. The method developed in this study provided robust estimates of functional and biophysical parameters for both airborne and synthetic EnMAP datasets. Cab and Vcmax estimates followed observed relationships with leaf nitrogen concentration; whereas m and predicted underlying water use efficiency showed expected relationships with discrimination of 13C isotope in leaves. Results prove that the inversion of coupled RTM-SVAT models against a combination of hyperspectral imagery (e.g., EnMAP), and time series of GPP and thermal irradiance provides reliable estimates of key functional parameters of vegetation that are robust to several sources of uncertainty. The forthcoming satellite hyperspectral missions combined with ecosystem station networks (e.g. Integrated Carbon Observation System (ICOS), NEON, FLUXNET, etc…), offers unique possibilities to characterize the spatiotemporal distribution of functional parameters relevant for terrestrial biosphere modeling.

scholarly journals A survey of proximal methods for monitoring leaf phenology in temperate deciduous forests

2021 ◽  
Vol 18 (11) ◽  
pp. 3391-3408
Author(s):  
Kamel Soudani ◽  
Nicolas Delpierre ◽  
Daniel Berveiller ◽  
Gabriel Hmimina ◽  
Jean-Yves Pontailler ◽  
...  
Keyword(s):  
Time Series ◽  
Interannual Variability ◽  
Leaf Senescence ◽  
Vegetation Index ◽  
Sigmoid Function ◽  
Spring Phenology ◽  
Area Index ◽  
Tree Phenology ◽  
Temperate Deciduous ◽  
Energy Exchanges

Abstract. Tree phenology is a major driver of forest–atmosphere mass and energy exchanges. Yet, tree phenology has rarely been monitored in a consistent way throughout the life of a flux-tower site. Here, we used seasonal time series of ground-based NDVI (Normalized Difference Vegetation Index), RGB camera GCC (greenness chromatic coordinate), broadband NDVI, LAI (leaf area index), fAPAR (fraction of absorbed photosynthetic active radiation), CC (canopy closure), fRvis (fraction of reflected radiation) and GPP (gross primary productivity) to predict six phenological markers detecting the start, middle and end of budburst and of leaf senescence in a temperate deciduous forest using an asymmetric double sigmoid function (ADS) fitted to the time series. We compared them to observations of budburst and leaf senescence achieved by field phenologists over a 13-year period. GCC, NDVI and CC captured the interannual variability of spring phenology very well (R2>0.80) and provided the best estimates of the observed budburst dates, with a mean absolute deviation (MAD) of less than 4 d. For the CC and GCC methods, mid-amplitude (50 %) threshold dates during spring phenological transition agreed well with the observed phenological dates. For the NDVI-based method, on average, the mean observed date coincides with the date when NDVI reaches 25 % of its amplitude of annual variation. For the other methods, MAD ranges from 6 to 17 d. The ADS method used to derive the phenological markers provides the most biased estimates for the GPP and GCC. During the leaf senescence stage, NDVI- and CC-derived dates correlated significantly with observed dates (R2=0.63 and 0.80 for NDVI and CC, respectively), with an MAD of less than 7 d. Our results show that proximal-sensing methods can be used to derive robust phenological metrics. They can be used to retrieve long-term phenological series at eddy covariance (EC) flux measurement sites and help interpret the interannual variability and trends of mass and energy exchanges.

scholarly journals Analogy-Based Crop Yield Forecasts Based on Temporal Similarity of Leaf Area Index

2021 ◽  
Vol 13 (16) ◽  
pp. 3069
Author(s):  
Yadong Liu ◽  
Junhwan Kim ◽  
David H. Fleisher ◽  
Kwang Soo Kim
Keyword(s):  
Time Series ◽  
Leaf Area Index ◽  
Leaf Area ◽  
Crop Yield ◽  
Satellite Data ◽  
Growing Season ◽  
Environmental Data ◽  
Area Index ◽  
Current Season ◽  
Wide Range

Seasonal forecasts of crop yield are important components for agricultural policy decisions and farmer planning. A wide range of input data are often needed to forecast crop yield in a region where sophisticated approaches such as machine learning and process-based models are used. This requires considerable effort for data preparation in addition to identifying data sources. Here, we propose a simpler approach called the Analogy Based Crop-yield (ABC) forecast scheme to make timely and accurate prediction of regional crop yield using a minimum set of inputs. In the ABC method, a growing season from a prior long-term period, e.g., 10 years, is first identified as analogous to the current season by the use of a similarity index based on the time series leaf area index (LAI) patterns. Crop yield in the given growing season is then forecasted using the weighted yield average reported in the analogous seasons for the area of interest. The ABC approach was used to predict corn and soybean yields in the Midwestern U.S. at the county level for the period of 2017–2019. The MOD15A2H, which is a satellite data product for LAI, was used to compile inputs. The mean absolute percentage error (MAPE) of crop yield forecasts was <10% for corn and soybean in each growing season when the time series of LAI from the day of year 89 to 209 was used as inputs to the ABC approach. The prediction error for the ABC approach was comparable to results from a deep neural network model that relied on soil and weather data as well as satellite data in a previous study. These results indicate that the ABC approach allowed for crop yield forecast with a lead-time of at least two months before harvest. In particular, the ABC scheme would be useful for regions where crop yield forecasts are limited by availability of reliable environmental data.

scholarly journals Identifying resting locations of a small elusive forest carnivore using a two-stage model accounting for GPS measurement error and hidden behavioral states

2021 ◽  
Vol 9 (1) ◽  
Author(s):  
Dalton J. Hance ◽  
Katie M. Moriarty ◽  
Bruce A. Hollen ◽  
Russell W. Perry
Keyword(s):  
Time Series ◽  
Measurement Error ◽  
Gaussian State ◽  
Parameter Estimates ◽  
Accelerometer Data ◽  
Two Stage ◽  
Location Data ◽  
Gps Measurement ◽  
Activity Measurements ◽  
Behavioral States

Abstract Background Studies of animal movement using location data are often faced with two challenges. First, time series of animal locations are likely to arise from multiple behavioral states (e.g., directed movement, resting) that cannot be observed directly. Second, location data can be affected by measurement error, including failed location fixes. Simultaneously addressing both problems in a single statistical model is analytically and computationally challenging. To both separate behavioral states and account for measurement error, we used a two-stage modeling approach to identify resting locations of fishers (Pekania pennanti) based on GPS and accelerometer data. Methods We developed a two-stage modelling approach to estimate when and where GPS-collared fishers were resting for 21 separate collar deployments on 9 individuals in southern Oregon. For each deployment, we first fit independent hidden Markov models (HMMs) to the time series of accelerometer-derived activity measurements and apparent step lengths to identify periods of movement and resting. Treating the state assignments as given, we next fit a set of linear Gaussian state space models (SSMs) to estimate the location of each resting event. Results Parameter estimates were similar across collar deployments. The HMMs successfully identified periods of resting and movement with posterior state assignment probabilities greater than 0.95 for 97% of all observations. On average, fishers were in the resting state 63% of the time. Rest events averaged 5 h (4.3 SD) and occurred most often at night. The SSMs allowed us to estimate the 95% credible ellipses with a median area of 0.12 ha for 3772 unique rest events. We identified 1176 geographically distinct rest locations; 13% of locations were used on > 1 occasion and 5% were used by > 1 fisher. Females and males traveled an average of 6.7 (3.5 SD) and 7.7 (6.8 SD) km/day, respectively. Conclusions We demonstrated that if auxiliary data are available (e.g., accelerometer data), a two-stage approach can successfully resolve both problems of latent behavioral states and GPS measurement error. Our relatively simple two-stage method is repeatable, computationally efficient, and yields directly interpretable estimates of resting site locations that can be used to guide conservation decisions.

Sign in / Sign up

Export Citation Format

Share Document