Surface Reflectance Time Series Research Articles

Boosting crop classification by hierarchically fusing satellite, rotational, and contextual data

Accurate early-season crop type classification is crucial for the crop production estimation and monitoring of agricultural parcels. However, the complexity of the plant growth patterns and their spatio-temporal variability present significant challenges. While current deep learning-based methods show promise in crop type classification from single- and multi-modal time series, most existing methods rely on a single modality, such as satellite optical remote sensing data or crop rotation patterns. We propose a novel approach to fuse multimodal information into a model for improved accuracy and robustness across multiple crop seasons and countries. The approach relies on three modalities used: remote sensing time series from Sentinel-2 and Landsat 8 observations, parcel crop rotation and local crop distribution. To evaluate our approach, we release a new annotated dataset of 7.4 million agricultural parcels in France (FR) and the Netherlands (NL). We associate each parcel with time-series of surface reflectance (Red and NIR) and biophysical variables (LAI, FAPAR). Additionally, we propose a new approach to automatically aggregate crop types into a hierarchical class structure for meaningful model evaluation and a novel data-augmentation technique for early-season classification. Performance of the multimodal approach was assessed at different aggregation levels in the semantic domain, yielding to various ranges of the number of classes spanning from 151 to 8 crop types or groups. It resulted in accuracy ranging from 91% to 95% for the NL dataset and from 85% to 89% for the FR dataset. Pre-training on a dataset improves transferability between countries, allowing for cross- domain and label prediction, and robustness of the performances in a few-shot setting from FR to NL, i.e., when the domain changes as per with significantly new labels. Our proposed approach outperforms comparable methods by enabling deep learning methods to use the often overlooked spatio-temporal context of parcels, resulting in increased precision and generalization capacity.

Using a Vegetation Index as a Proxy for Reliability in Surface Reflectance Time Series Reconstruction (RTSR)

Time series of optical remote sensing data are instrumental for monitoring vegetation dynamics, but are hampered by missing or noisy observations due to varying atmospheric conditions. Reconstruction methods have been proposed, most of which focus on time series of a single vegetation index. Under the assumption that relatively high vegetation index values can be considered as trustworthy, a successful approach is to adjust the smoothed value to the upper envelope of the time series. However, this assumption does not hold for surface reflectance in general. Clouds and cloud shadows result in, respectively, high and low values in the visible and near infrared part of the electromagnetic spectrum. A novel spectral Reflectance Time Series Reconstruction (RTSR) method is proposed. Smoothed values of surface reflectance values are adjusted to approach the trustworthy observations, using a vegetation index as a proxy for reliability. The Savitzky–Golay filter was used as the smoothing algorithm here, but different filters can be used as well. The RTSR was evaluated on 100 sites in Europe, with a focus on agriculture fields. Its potential was shown using different criteria, including smoothness and the ability to retain trustworthy observations in the original time series with RMSE values in the order of 0.01 to 0.03 in terms of surface reflectance.

Predicting plant biomass and species richness in temperate grasslands across regions, time, and land management with remote sensing and deep learning

Spatial predictions of biomass production and biodiversity at regional scale in grasslands are critical to evaluate the effects of management practices across environmental gradients. New generations of remote sensing sensors and machine learning approaches can predict these grassland characteristics with varying accuracy. However, such studies frequently fail to cover a sufficiently broad range of environmental conditions, and their prediction models are often case-specific. To address this gap, we have modelled above-ground biomass and species richness in 150 spatially independent grassland plots of three geographical regions in Germany. These regions follow a North-South climate gradient and differ in soil types, topography, elevation, climatic conditions, historical contexts, and management intensities. The predictors tested in this study are Sentinel-1 backscatter, Sentinel-2 time series of surface reflectance along with derived vegetation indices and Rao's Q, and a set of topoedaphic variables. We compared the performance of a feed-forward deep neural network (DNN) with a random forest (RF) regression algorithm. The DNN achieved the best estimations of biomass (r2 = 0.45) when trained with Sentinel-2 surface reflectance only. Moreover, the DNN showed a higher generalizability than RF during spatial cross-validations (i.e., calibrating and validating in different regions, r2 = 0.38 vs. 0.26). Species richness predictions by both algorithms improved when the full time series of Sentinel-2 surface reflectance values were used (highest r2 = 0.42 achieved by the DNN), but both performed poorly during spatial cross-validations. Overall, the DNN-based models were more robust than RF models, showed a lower bias and lower systematic error, and required fewer inputs. Explainability analysis indicated that red-edge and near infrared information from May and October was the most relevant to predict species richness. This study presents an important step forward in generating robust spatially explicit predictions of grassland attributes and biodiversity variables across large areas, environmental gradients, and phenological stages.

Reconstruction of Sentinel-2 Image Time Series Using Google Earth Engine

Sentinel-2 NDVI and surface reflectance time series have been widely used in various geoscience research, but the data is deteriorated or missing due to the cloud contamination, so it is necessary to reconstruct the Sentinel-2 NDVI and surface reflectance time series. At present, there are few studies on reconstructing the Sentinel-2 NDVI or surface reflectance time series, and these existing reconstruction methods have some shortcomings. We proposed a new method to reconstruct the Sentinel-2 NDVI and surface reflectance time series using the penalized least-square regression based on discrete cosine transform (DCT-PLS) method. This method iteratively identifies cloud-contaminated NDVI over NDVI time series from the Sentinel-2 surface reflectance data by adjusting the weights. The NDVI and surface reflectance time series are then reconstructed from cloud-free NDVI and surface reflectance using the adjusted weights as constraints. We have made some improvements to the DCT-PLS method. First, the traditional discrete cosine transformation (DCT) in the DCT-PLS method is matrix generated from discrete and equally spaced data, we reconfigured the DCT formulas to adapt for irregular interval time series, and optimized the control parameters N and s according to the typical vegetation samples in China. Second, the DCT-PLS method was deployed in the Google Earth Engine (GEE) platform for the efficiency and convenience of data users. We used the DCT-PLS method to reconstruct the Sentinel-2 NDVI time series and surface reflectance time series in the blue, green, red, and near infrared (NIR) bands in typical vegetation samples and the Zhangjiakou and Hangzhou study area. We found that this method performed better than the SG filter method in reconstructing the NDVI time series, and can identify and reconstruct the contaminated NDVI as well as surface reflectance with low root mean square error (RMSE) and high coefficient of determination (R2). However, in cases of a long range of cloud contamination, or above water surface, it may be necessary to increase the control parameter s for a more stable performance. The GEE code is freely available online and the link is in the conclusions of this article, researchers are welcome to use this method to generate cloudless Sentinel-2 NDVI and surface reflectance time series with 10 m spatial resolution, which is convenient for landcover classification and many other types of research.

Development of the GLASS 250-m leaf area index product (version 6) from MODIS data using the bidirectional LSTM deep learning model

Leaf area index (LAI) is a terrestrial essential climate variable that is required in a variety of ecosystem and climate models. The Global LAnd Surface Satellite (GLASS) LAI product has been widely used, but its current version (V5) from Moderate Resolution Imaging Spectroradiometer (MODIS) data has several limitations, such as frequent temporal fluctuation, large data gaps, high dependence on the quality of surface reflectance, and low computational efficiency. To address these issues, this paper presents a deep learning model to generate a new version of the LAI product (V6) at 250-m resolution from MODIS data from 2000 onward. Unlike most existing algorithms that estimate one LAI value at one time for each pixel, this model estimates LAI for 2 years simultaneously. Three widely used LAI products (MODIS C6, GLASS V5, and PROBA-V V1) are used to generate global representative time-series LAI training samples using K-means clustering analysis and least difference criteria. We explore four machine learning models, the general regression neural network (GRNN), long short-term memory (LSTM), gated recurrent unit (GRU), and Bidirectional LSTM (Bi-LSTM), and identify Bi-LSTM as the best model for product generation. This new product is directly validated using 79 high-resolution LAI reference maps from three in situ observation networks. The results show that GLASS V6 LAI achieves higher accuracy, with a root mean square (RMSE) of 0.92 at 250 m and 0.86 at 500 m, while the RMSE is 0.98 for PROBA-V at 300 m, 1.08 for GLASS V5, and 0.95 for MODIS C6 both at 500 m. Spatial and temporal consistency analyses also demonstrate that the GLASS V6 LAI product is more spatiotemporally continuous and has higher quality in terms of presenting more realistic temporal LAI dynamics when the surface reflectance is absent for a long period owing to persistent cloud/aerosol contaminations. The results indicate that the new Bi-LSTM deep learning model runs significantly faster than the GLASS V5 algorithm, avoids the reconstruction of surface reflectance data, and is resistant to the noises (cloud and snow contamination) or missing values contained in surface reflectance than other methods, as the Bi-LSTM can effectively extract information across the entire time series of surface reflectance rather than a single time point. To our knowledge, this is the first global time-series LAI product at the 250-m spatial resolution that is freely available to the public (www.geodata.cn and www.glass.umd.edu).

A novel regression method for harmonic analysis of time series

Harmonic analysis of time series is an important technique to reveal seasonal land surface dynamics using remote sensing information. However, frequency selection in the harmonic analysis is often difficult because high-frequency components are useful for delineating seasonal dynamics but sensitive to noise and gaps in time series. On the other hand, it is challenging to obtain temporally continuous satellite data with high quality because of atmospheric contamination. We developed a novel regression method named Harmonic Adaptive Penalty Operator (HAPO) for harmonic analysis of unevenly distributed time series. We introduced a new penalty function to minimize unexpected fluctuations in the model, which can substantially reduce the overfitting issue of regression in time series with temporal gaps. Specifically, the new penalty function minimizes the length of the model curve and the value range difference between the model and time series observations. We compared HAPO with three widely used regression methods (OLS: Ordinary Least Squares; LASSO: Least Absolute Shrinkage and Selection Operator; and Ridge) with different scenarios using Landsat time series data across the United States. First, we evaluated methods using Landsat surface reflectance time series within a single year. HAPO showed small and consistent monthly Root Mean Square Deviation (RMSD) values, in which most of the time RMSD values of predicted reflectance were less than 0.04. More importantly, HAPO showed consistent and less bias given varying density and irregularity of time series. Second, we evaluated methods using multi-year time series and the result suggested that HAPO was a better predictor of relatively short time series (less than4 years) with steady small RMSD values. When a longer time series (≥4 years) was used, all four methods disclosed similar RMSD values, but HAPO outperformed other three methods when there were temporal gaps. Last, we preliminarily tested how regression methods affected change detection and classification accuracy. HAPO showed the highest change detection accuracy of all tests in terms of F1 score when using the change threshold of 0.9999. In classification, HAPO produced the highest accuracy for short time series segments (one- or two-year time series). In contrast, all methods reached similar accuracy for 5-year time series. These results suggest that for areas that have large seasonal observation gaps or for time series that have less than 4 years records, HAPO can provide more consistent and accurate analytical results than other regression methods for harmonic analysis of time series.

Antarctic biological soil crusts surface reflectance patterns from landsat and sentinel-2 images.

The remote sensing techniques must be used to obtain long-term information in remote areas, like the Antarctic continent, to monitor the environmental productivity and its changes. The aim of this work was to analyze the surface reflectance profile patterns for the Antarctic biological soil crusts (algae, lichens, and mosses) in an area of Nelson Island (South Shetland Islands, maritime Antarctic), calculated from Landsat and Sentinel-2 images to identify its similarities and differences due to targets, sensors and acquired date. The surface reflectance values for Antarctic biological soil crusts are similar for those observed for biological soil crusts in other Earth extreme environments, like deserts. In Landsat images, the differences among biological soil crusts surface reflectance were identified at visible and near-infrared wavelengths and for Sentinel-2 images, the differences occur at visible, red-edge and shortwave infrared wavelengths, showing the feasibility of using surface reflectance products to identify these different crusts, despite its inherent pixel spectral mixture. Long-term biophysical parameters from such crusts as retrieved from orbital data is not possible due to very low cloud-free images over the Antarctic, which prevents building a consistent surface reflectance time-series which covers all biological soil crusts growth season.

Evaluation of Orbital Drift Effect on Proba-V Surface Reflectances Time Series

Multi-temporal consistency of space-borne observations is an essential requirement for studying inter-annual changes and trends of satellite-derived biophysical products. The Proba-V mission, launched in 2013, was designed to ensure the continuity of the SPOT-VEGETATION long-term data record of global daily observations for land applications. The suitability of Proba-V to provide a temporally consistent data record is, however, potentially jeopardized by the orbital drift effect, which is known to induce spurious trends in time series. The aim of this paper is therefore to evaluate, for the first time, the orbital drift effect on Proba-V surface reflectance time series at 1 km resolution. In order to reliably identify such an effect, a two-fold approach is adopted. A simulation study is first defined to predict the temporal anomalies induced by the drifting illumination conditions. The numerical simulations are used as a benchmark to predict the impact of the drift for a range of sun-viewing angles. Real observations are then analyzed over a large set of land sites, globally spread and spanning a wide range of surface and environmental conditions. The surface anisotropy is characterized using the Ross-Thick Li-Sparse Reciprocal (RTLSR) Bidirectional Reflectance Distribution Function (BRDF) model. Both the simulation and the analysis of real observations consistently show that the orbital drift induces distinct and opposite trends in the two sides of the sensor across-track swath. Particularly, a positive drift is estimated in backward and a negative one in the forward scattering direction. When observations from all angular conditions are retained, these opposite trends largely compensate, with no remaining statistically significant drifts in time series of surface reflectances or Normalized Difference Vegetation Index (NDVI). As such, the Proba-V archive at 1 km resolution can be reliably used for inter-annual vegetation studies.

Empirical model for forecasting sugarcane yield on a local scale in Brazil using Landsat imagery and random forest algorithm

Sugarcane plays an important role in food and energy production in Brazil and worldwide. The large availability of satellite sensors and advanced techniques for processing data have improved the forecasting sugarcane yield on a local and global scale, but more work is needed on exploiting the synergy between remote sensing, meteorological and agronomic data. In this study, we combined such data sources to forecast sugarcane yield using a random forest (RF) algorithm on an extensive area of 50,000 ha, over four years. Images from Landsat satellites were processed to time series of surface reflectance and spectral indices. The approach focused on the development of predictive models which only used data acquired and accessible several months before the harvest. First, three RF models were calibrated with different predictors to forecast the sugarcane yield at harvest: using Landsat satellite images and meteorological data (RF1); agronomic and meteorological data (RF2); a combination of Landsat satellite images, agronomic and meteorological data (RF3). As a comparison, we also tested the influence of including knowledge on the future harvest date in the models RF2 and RF3 (RF4 and RF5). The average values of R2 for RF1, RF2, and RF3 were 0.66, 0.50 and 0.74, respectively. The model with the highest values of R2 (RF3) had a Root Mean Square Error (RMSE) of 9.9 ton ha−1 on yield forecast, approximately 15% of the yield average. Including the harvest date improved the RF2 and RF3 models to reach R2 = 0.69 and RMSE = 10.8 ton ha−1 for RF4, and R2 = 0.76 and RMSE of 9.4 ton ha−1 for RF5. A blind forecasting test for the 2016 yields showed similar prediction than the forecast made by in situ field expertise. This result has the potential to assist management of sugarcane production.

Improving Time Series Reconstruction by Fixing Invalid Values and its Fidelity Evaluation

MODIS time series data have been widely used in the research of regional and global ecosystems and climate change. For vegetation monitoring, vegetation indices such as NDVI (normalized difference vegetation index), EVI (enhanced vegetation index) and NBR (normalized burn ratio), are usually derived from MODIS reflectance data. However, noise usually makes it difficult to generate reliable time series of vegetation indices. Although some methods have been developed for reconstructing NDVI time series data, they still suffer from some limitations. First, there is no reliable approach for detecting and dealing with low-quality data, resulting in poor outcomes. Second, no effective evaluation of the fidelity of the corrected data to the original data has been discussed. For these reasons, we developed a new time series reconstruction approach, named Fixing Invalid Value (FIV) method. The proposed method assumes that the noise in surface reflectance data stems from invalid data, such as clouds, ice, and missing values. The FIV method first uses the spatially and temporally neighboring pixels to estimate the invalid values and then applies morphology operations to remove the residual noise. Finally, the Savitzky-Golay (S-G) filter is employed to generate the final results. The FIV method is tested on 8-day composite MODIS surface reflectance time series data from 2001 to 2012 in Jiangxi and Fujian provinces, China. The results show that the FIV method outperforms the conventional S-G filter and the HANTS method both in terms of visual inspection and quantitative evaluation. Furthermore, the fidelity evaluation reveals that the proposed FIV method produces high-quality time series data under all weather conditions.

A patch-based algorithm for global and daily burned area mapping

Increasing availability of dense time series of moderate spatial resolution satellite data for mapping global burned areas calls for mapping algorithms designed to easily integrate data at different spatial and temporal resolutions, irrespective of particular grid constraints. In this paper, we describe a novel hybrid approach for global burned area mapping that combines active fire data and time series of surface reflectance using graphs, which provide a flexible and efficient way of extracting spatiotemporal consistent patches.Our approach has three main steps. Firstly, we analyze burn-sensitive vegetation index time series to determine for each location a set of events, which are the dates for which the spectral-temporal signal indicates the possibility of a burn. Secondly, we explore the spatiotemporal distribution of all events and active fires to determine a subset of events with strong evidence of corresponding to burned areas. Those events are used as positive occurrences for training a one-class maximum entropy classifier and obtain, for each candidate event, a likelihood of it actually corresponding to a burn. Finally, we build a graph that combines all previous information, from which we extract spatiotemporal patches of densely connected events. Patches with strong evidence of burning determine the burned area map at any given time period.This research is part of the European Space Agency's Climate Change Initiative (ESA-CCI) and aims ultimately at generating Sentinel-3 daily global burned area products at 500 m spatial resolution. Towards that end, we test our approach with spatially and spectrally similar MODIS gridded surface reflectance (MOD/MYD09GA), as well as non-gridded active fire (MCD14ML) 2008 data and CCI global land cover maps. Using 105 independent Landsat fire reference perimeters to validate global results, we show that our algorithm applied to MOD/MYD09GA data (PT-M09) has very similar accuracy (44%) measured by the Dice coefficient compared with MCD64A1 v006 (45%). Moreover, PT-M09 exhibits a higher commission error but a lower omission error than MCD64A1. Due to their coarse resolution, this kind of product cannot capture very small burn areas. The bias relative to the reference burned area indicates that the algorithm presented in the current study underestimates burned area by 14% of the area actually burned according to reference data, which is lower than the underestimation in MCD64A1 (28%).We also analyzed the temporal accuracy of the patch based algorithm and concluded that the average time difference to active fires detection is 3.43 days, which is similar to MCD64A1 (3.75 days). Finally, we performed a sensitivity analysis which shows that total mapped burned areas varies only 3% when the main algorithm threshold that separates “burned” and “unburned” patches varies from quantile 45% to 55%.

Continuous monitoring of land disturbance based on Landsat time series

We developed a new algorithm for COntinuous monitoring of Land Disturbance (COLD) using Landsat time series. COLD can detect many kinds of land disturbance continuously as new images are collected and provide historical land disturbance maps retrospectively. To better detect land disturbance, we tested different kinds of input data and explored many time series analysis techniques. We have several major observations as follows. First, time series of surface reflectance provides much better detection results than time series of Top-Of-Atmosphere (TOA) reflectance, and with some adjustments to the temporal density, time series from Landsat Analysis Ready Data (ARD) is better than it is from the same Landsat scene. Second, the combined use of spectral bands is always better than using a single spectral band or index, and if all the essential spectral bands have been employed, the inclusion of other indices does not further improve the algorithm performance. Third, the remaining outliers in the time series can be removed based on their deviation from model predicted values based on probability-based thresholds derived from normal or chi-squared distributions. Fourth, model initialization is pivotal for monitoring land disturbance, and a good initialization stability test can influence algorithm performance substantially. Fifth, time series model estimation with eight coefficients model, updated for every single observation, based on all available clear observations achieves the best result. Sixth, a change probability of 0.99 (chi-squared distribution) with six consecutive anomaly observations and a mean included angle < 45° to confirm a change provide the best results, and the combined use of temporally-adjusted Root Mean Square Error (RMSE) and minimum RMSE is recommended. Finally, spectral changes (or “breaks”) contributed from vegetation regrowth should be excluded from land disturbance maps. The COLD algorithm was developed and calibrated based on all these lessons learned above. The accuracy assessment shows that COLD results were accurate for detecting land disturbance, with an omission error of 27% and a commission error of 28%.

Improving the AVHRR Long Term Data Record BRDF Correction

The Long Term Data Record (LTDR) project has the goal of developing a quality and consistent surface reflectance product from coarse resolution optical sensors. This paper focuses on the Advanced Very High Resolution Radiometer (AVHRR) part of the record, using the Moderate Resolution Imaging Spectrometer (MODIS) instrument as a reference. When a surface reflectance time series is acquired from satellites with variable observation geometry, the directional variation generates an apparent noise which can be corrected by modeling the bidirectional reflectance distribution function (BRDF). The VJB (Vermote, Justice and Bréon, 2009) method estimates a target’s BRDF shape using 5 years of observation and corrects for directional effects maintaining the high temporal resolution of the measurement using the instantaneous Normalized Difference Vegetation Index (NDVI). The method was originally established on MODIS data but its viability and optimization for AVHRR data have not been fully explored. In this study we analyze different approaches to find the most robust way of applying the VJB correction to AVHRR data, considering that high noise in the red band (B1) caused by atmospheric effect makes the VJB method unstable. Firstly, our results show that for coarse spatial resolution, where the vegetation dynamics of the target don’t change significantly, deriving BRDF parameters from 15+ years of observations reduces the average noise by up to 7% in the Near Infrared (NIR) band and 6% in the NDVI, in comparison to using 3-year windows. Secondly, we find that the VJB method can be modified for AVHRR data to improve the robustness of the correction parameters and decrease the noise by an extra 8% and 9% in the red and NIR bands with respect to using the classical VJB inversion. We do this by using the Stable method, which obtains the volumetric BRDF parameter (V) based on its NDVI dependency, and then obtains the geometric BRDF parameter (R) through the inversion of just one parameter.

Toward Landsat and Sentinel-2 BRDF Normalization and Albedo Estimation: A Case Study in the Peruvian Amazon Forest

The Amazon forest has been the focus of study by the science community during the last few decades. Remote sensing data analysis is the only way to study such a large geographical extent during an extended period of time. Since the launch in 2015 of Sentinel 2 and its increase in temporal resolution through the combination with Landsat sensors, a strong emphasis has been put on exploiting these data. Though these satellites provide near nadir observations, surface reflectance time series are affected by illumination variability throughout the year. These effects can be corrected using a Bidirectional Reflectance Distribution Function (BRDF) model. Franch et al. (2014a) developed a methodology to derive Landsat surface albedo and BRDF. It is based on the BRDF parameters from the MODerate Resolution Imaging Spectroradiometer (MODIS) which are disaggregated at Landsat spatial resolution (30 m). In this work, we apply the Franch et al. (2014a) method to normalize the surface reflectance for BRDF effects using the NASA's Harmonized Landsat Sentinel-2 (HLS) product. We apply this method to the Tambopata region in Peru from 2013 to 2017 and validate it using ground-based albedometer measurements. The results show that the near infrared reflectance can increase up to 0.06 (20%) for low solar angles while the impact on the red range and the NDVI is minor (<0.01). The evaluation of the surface albedo against field measurements shows an error of 0.01.

Simulating arbitrary hyperspectral bandsets from multispectral observations via a generic Earth Observation-Land Data Assimilation System (EO-LDAS)

This paper presents results of using multi-sensor and multi-angular constraints in the generic Earth Observation-Land Data Assimilation System (EO-LDAS) for reproducing arbitrary bandsets of hyperspectral reflectance at the top-of-canopy (TOC) level by merging observations from multispectral sensors with different spectral characteristics. This is demonstrated by combining Multi-angle Imaging Spectroradiometer (MISR) and Landsat Enhanced Thematic Mapper Plus (ETM+) data to simulate the Compact High Resolution Imaging Spectrometer CHRIS/PROBA hyperspectral signal over an agricultural test site, in Barrax, Spain. However, the method can be more generally applied to any combination of spectral data, providing a tool for merging EO data to any arbitrary hyperspectral bandset.Comparisons are presented using both synthetic and observed MISR and Landsat data, and retrieving surface biophysical properties. We find that when using simulated MISR and Landsat data, the CHRIS/PROBA hyperspectral signal is reproduced with RMSE 0.0001–0.04. LAI is retrieved with r2 from 0.97 to 0.99 and RMSE of from 0.21 to 0.38. The results based on observed MISR and Landsat data have lower performances, with RMSE for the reproduced CHRIS/PROBA hyperspectral signal varying from 0.007 to 0.2. LAI is retrievedwith r2 from 0.7 to 0.9 and RMSE from 0.7 to 1.4. We found that for the data considered here the main spectral variations in the visible and near infrared regions can be described by a limited number of parameters (3–4) that can be estimated from multispectral information. Results show that the method can be used to simulate arbitrary bandsets, which will be of importance to any application which requires combining new and existing streams of new EO data in the optical domain, particularly intercalibration of EO satellites in order to get continuous time series of surface reflectance, across programmes and sensors of different designs.

A Cubesat enabled Spatio-Temporal Enhancement Method (CESTEM) utilizing Planet, Landsat and MODIS data

Satellite sensing in the visible to near-infrared (VNIR) domain has been the backbone of land surface monitoring and characterization for more than four decades. However, a limitation of conventional single-sensor satellite missions is their limited capacity to observe land surface dynamics at the very high spatial and temporal resolutions demanded by a wide range of applications. One solution to this spatio-temporal divide is an observation strategy based on the CubeSat standard, which facilitates constellations of small, inexpensive satellites. Repeatable near-daily image capture in RGB and near-infrared (NIR) bands at 3–4 m resolution has recently become available via a constellation of >130 CubeSats operated commercially by Planet. While the observing capacity afforded by this system is unprecedented, the relatively low radiometric quality and cross-sensor inconsistencies represent key challenges in the realization of their full potential as a game changer in Earth observation. To address this issue, we developed a Cubesat Enabled Spatio-Temporal Enhancement Method (CESTEM) that uses a multi-scale machine-learning technique to correct for radiometric inconsistencies between CubeSat acquisitions. The CESTEM produces Landsat 8 consistent atmospherically corrected surface reflectances in blue, green, red, and NIR bands, but at the spatial scale and temporal frequency of the CubeSat observations. An application of CESTEM over an agricultural dryland system in Saudi Arabia demonstrated CubeSat-based reproduction of Landsat 8 consistent VNIR data with an overall relative mean absolute deviation of 1.6% or better, even when the Landsat 8 and CubeSat acquisitions were temporally displaced by >32 days. The consistently high retrieval accuracies were achieved using a multi-scale target sampling scheme that draws Landsat 8 reference data from a series of scenes by using MODIS-consistent surface reflectance time series to quantify relative changes in Landsat-scale reflectances over given Landsat-CubeSat acquisition timespans. With the observing potential of Planet's CubeSats approaching daily nadir-pointing land surface imaging of the entire Earth, CESTEM offers the capacity to produce daily Landsat 8 consistent VNIR imagery with a factor of 10 increase in spatial resolution and with the radiometric quality of actual Landsat 8 observations. Realization of this unprecedented Earth observing capacity has far reaching implications for the monitoring and characterization of terrestrial systems at the precision scale.

Color and temperature of the crater lakes at Kelimutu volcano through time

We investigated the color and temperature of three volcanic crater lakes that co-exist at Kelimutu volcano (Indonesia) using ~ 30 years of Landsat data. These satellite data were obtained through Google Earth Engine. Time series of surface reflectance (visible wavelengths) and brightness temperature above background (thermal infrared wavelengths) were calculated. Color was defined in the RGB (red-green-blue) and HSV (hue-saturation-value) color spaces, and we introduce a visualization concept called “hue stretch” to consistently represent hue through time. These parameters display long-term trends, seasonal cycles and short duration bursts of activity at the lakes. We demonstrate that the color of the lakes are related over a period of months to years and discovered a previously unreported but significant episode around 1997, which included large agglomerations of floating elemental sulfur. Globally speaking, these techniques could reveal trends at any of the ~ 100 crater lakes on active volcanoes. Furthermore, they could apply to any target whose color changes through time (e.g., forests, crops, and non-volcanic water bodies). We have open-sourced the code necessary to perform these analyses.

Reconstruction of Long-Term Temporally Continuous NDVI and Surface Reflectance From AVHRR Data

Advanced very high resolution radiometer (AVHRR) data provide the longest available time series of global satellite observations and have been extensively used. The Land Long-Term Data Record (LTDR) project has generated daily surface reflectance and normalized difference vegetation index (NDVI) products from AVHRR. However, residual cloud and aerosol contamination in the LTDR AVHRR surface reflectance and NDVI products significantly limits their applications and results in temporal and spatial inconsistencies in subsequent downstream products. Based on the LTDR AVHRR surface reflectance, a temporally continuous vegetation indices-based land-surface reflectance reconstruction (VIRR) method was refined in this study to generate Global LAnd Surface Satellite (GLASS) AVHRR NDVI and surface reflectance products from 1982 to 2015. The daily LTDR AVHRR surface reflectance data were first aggregated into eight-day intervals. The aggregated surface reflectance data were used to calculate NDVI, and a robust smoothing algorithm was used to reconstruct continuous and smooth NDVI upper envelopes, which were used to identify cloud-contaminated surface reflectance values. Then the surface reflectance time series was reconstructed from cloud-free surface reflectance values by incorporating the upper envelopes of the NDVI time series as constraints. The results show that the refined VIRR method successfully removes NDVI and surface reflectance values contaminated by clouds and can reconstruct temporally continuous NDVI and land-surface reflectance time series. Comparison of the GLASS AVHRR NDVI product with the third-generation Global Inventory Monitoring and Modeling System (GIMMS3g) and the moderate resolution imaging spectroradiometer (MODIS) NDVI products indicates that these NDVI products exhibit similar spatial patterns, but the GIMMS3g NDVI values were clearly higher than the GLASS AVHRR and MODIS NDVI values in tropical forest regions and the 50°N−60°N latitude band, particularly in July. Comparisons with the MODIS NDVI values over the BELMANIP (Benchmark Land Multisite Analysis and Intercomparison of Products) sites demonstrate that the GLASS AVHRR NDVI product provides better performance (RMSE = 0.1007 and Bias = 0.0518) than the GIMMS3g NDVI product (RMSE = 0.1288 and Bias = 0.0852). The temporal profiles of all these NDVI products exhibited consistent seasonal variations, but the temporal smoothness of the GLASS AVHRR NDVI product was superior to that of the GIMMS3g and MODIS NDVI products. The GLASS AVHRR and GIMMS3g NDVI products show consistent trends in most situations, but the trends of the GLASS AVHRR NDVI product were slightly more pronounced than those of the GIMMS3g NDVI product for each biome type. Comparison of the GLASS AVHRR surface reflectance product with MODIS surface reflectance product indicates the GLASS AVHRR and MODIS surface reflectance showed similar seasonal and interannual variations and the GLASS AVHRR surface reflectance was in good agreement with the MODIS surface reflectance, especially in the red band.

A 30+ year AVHRR Land Surface Reflectance Climate Data Record and its application to wheat yield monitoring.

The Advanced Very High Resolution Radiometer (AVHRR) sensor provides a unique global remote sensing dataset that ranges from the 1980’s to the present. Over the years, several efforts have been made on the calibration of the different instruments to establish a consistent land surface reflectance time-series and to augment the AVHRR data record with data from other sensors such as the Moderate Resolution Imaging Spectroradiometer (MODIS). In this paper, we present a summary of all the corrections applied to the AVHRR Surface Reflectance and NDVI Version 4 Product, developed in the framework of the National Oceanic and Atmospheric Administration (NOAA) Climate Data Record (CDR) program. These corrections result from assessment of the geo-location, improvement of the cloud masking and calibration monitoring. Additionally, we evaluate the performance of the surface reflectance over the AERONET sites by a cross-comparison with MODIS, which is an already validated product, and evaluation of a downstream Leaf Area Index (LAI) product. We demonstrate the utility of this long time-series by estimating the winter wheat yield over the USA. The methods developed by [1] and [2] are applied to both the MODIS and AVHRR data. Comparison of the results from both sensors during the MODIS-era shows the consistency of the dataset with similar errors of 10%. When applying the methods to AVHRR historical data from the 1980’s, the results have errors equivalent to those derived from MODIS.

Reconstruction of Satellite-Retrieved Land-Surface Reflectance Based on Temporally-Continuous Vegetation Indices

Land-surface reflectance, estimated from satellite observations through atmospheric corrections, is an essential parameter for further retrieval of various high level land-surface parameters, such as leaf area index (LAI), fraction of absorbed photosynthetically active radiation (FAPAR), and surface albedo. Although great efforts have been made, land-surface reflectance products still contain considerable noise caused by, e.g., cloud or mixed-cloud pixels, which results in temporal and spatial inconsistencies in subsequent downstream products. In this study, a new method is developed to remove the residual clouds in the Moderate Resolution Imaging Spectroradiometer (MODIS) land-surface reflectance product and reconstruct time series of surface reflectance for the red, near infrared (NIR), and shortwave infrared (SWIR) bands. A smoothing method is introduced to calculate upper envelopes of vegetation indices (VIs) from the surface reflectance data and the cloud contaminated reflectance data are identified using the time series VIs and the upper envelopes of the time series VIs. Surface reflectance was then reconstructed according to cloud-free surface reflectance by incorporating the upper envelopes of the time series VIs as constraint conditions. The method was applied to reconstruct time series of surface reflectance from MODIS/TERRA surface reflectance product (MOD09A1). Temporal consistency analysis indicates that the new method can reconstruct temporally-continuous time series of land-surface reflectance. Comparisons with cloud-free MODIS/AQUA surface reflectance product (MYD09A1) over the BELMANIP (Benchmark Land Multisite Analysis and Intercomparison of Products) sites in 2003 demonstrate that the new method provides better performance for the red band (R2 = 0.8606 and RMSE = 0.0366) and NIR band (R2 = 0.6934 and RMSE = 0.0519), than the time series cloud detection (TSCD) algorithm (R2 = 0.5811 and RMSE = 0.0649; and R2 = 0.5005 and RMSE = 0.0675, respectively).