Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Sparse estimation”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 397 records · Page 22

An Adaptive Sparse-Grid High-Order Stochastic Collocation Method for Bayesian Inference in Groundwater Reactive Transport Modeling

Although Bayesian analysis has become vital to the quantification of prediction uncertainty in groundwater modeling, its application has been hindered due to the computational cost associated with numerous model executions needed for exploring the posterior probability density function (PPDF) of model parameters. This is particularly the case when the PPDF is estimated using Markov Chain Monte Carlo (MCMC) sampling. In this study, we develop a new approach that improves computational efficiency of Bayesian inference by constructing a surrogate system based on an adaptive sparse-grid high-order stochastic collocation (aSG-hSC) method. Unlike previous works using first-order hierarchical basis, we utilize a compactly supported higher-order hierar- chical basis to construct the surrogate system, resulting in a significant reduction in the number of computational simulations required. In addition, we use hierarchical surplus as an error indi- cator to determine adaptive sparse grids. This allows local refinement in the uncertain domain and/or anisotropic detection with respect to the random model parameters, which further improves computational efficiency. Finally, we incorporate a global optimization technique and propose an iterative algorithm for building the surrogate system for the PPDF with multiple significant modes. Once the surrogate system is determined, the PPDF can be evaluated by sampling the surrogate system directly with very little computational cost. The developed method is evaluated first using a simple analytical density function with multiple modes and then using two synthetic groundwater reactive transport models. The groundwater models represent different levels of complexity; the first example involves coupled linear reactions and the second example simulates nonlinear ura- nium surface complexation. The results show that the aSG-hSC is an effective and efficient tool for Bayesian inference in groundwater modeling in comparison with conventional MCMC sim- ulations. The computational efficiency is expected to be more beneficial to more computational expensive groundwater problems.

Zhang, Guannan↗

The Global Fire Atlas of Individual Fire Size, Duration, Speed and Direction

Natural and human-ignited fires affect all major biomes, altering ecosystem structure, biogeochemical cycles and atmospheric composition. Satellite observations provide global data on spatiotemporal patterns of biomass burning and evidence for the rapid changes in global fire activity in response to land management and climate. Satellite imagery also provides detailed information on the daily or sub-daily position of fires that can be used to understand the dynamics of individual fires. The Global Fire Atlas is a new global dataset that tracks the dynamics of individual fires to determine the timing and location of ignitions, fire size and duration, and daily expansion, fire line length, speed, and direction of spread. Here, we present the underlying methodology and Global Fire Atlas results for 2003–2016 derived from daily moderate-resolution (500 m) Collection 6 MCD64A1 burned-area data. The algorithm identified 13.3 million individual fires over the study period, and estimated fire perimeters were in good agreement with independent data for the continental United States. A small number of large fires dominated sparsely populated arid and boreal ecosystems, while burned area in agricultural and other human-dominated landscapes was driven by high ignition densities that resulted in numerous smaller fires. Long-duration fires in boreal regions and natural landscapes in the humid tropics suggest that fire season length exerts a strong control on fire size and total burned area in these areas. In arid ecosystems with low fuel densities, high fire spread rates resulted in large, short-duration fires that quickly consumed available fuels. Importantly, multiday fires contributed the majority of burned area in all biomass burning regions. A first analysis of the largest, longest and fastest fires that occurred around the world revealed coherent regional patterns of extreme fires driven by large-scale climate forcing. Global Fire Atlas data are publicly available through http://www.globalfiredata.org (last access: 9 August 2018) and https://doi.org/10.3334/ORNLDAAC/1642, and individual fire information and summary data products provide new information for benchmarking fire models within ecosystem and Earth system models, understanding vegetation–fire feedbacks, improving global emissions estimates, and characterizing the changing role of fire in the Earth system.

global fire atlas↗

The bioclimatic extent and pattern of the cold edge of the boreal forest: the circumpolar taiga-tundra ecotone

Current configurations of forest structure at the cold edge of the boreal may help understand the future of ecosystem functioning in high northern latitudes. The circumpolar biome boundary at the boreal (taiga) forest and tundra interface is an ecological transition zone (taiga-tundra ecotone; TTE) experiencing changes that affect its forest structure. We accounted for the TTE's horizontal forest structure with an estimate of its extent and pattern as represented by tree canopy cover (TCC). We quantified TCC patterns with an algorithm that describes its spatial gradient, and summarized landscape patterns of structure to represent heterogeneity, capturing abrupt, diffuse, and uniform forest at mesoscales. We used these landscape patterns to constrain the spatial extent of sparse and open canopy forest, and non-forest (forest-adjacent) edge that defines the TTE extent. The resulting map of the TTE extent is based on forest structure spatial patterns resolved at 30 m, highlights structural variability across landscapes, and helps distinguish tundra from boreal domains. We classified 14 594 landscapes as those associated with the TTE within a circumpolar bioclimatic envelope (11.575 million km2), where 44.83% of the area of these landscapes were forest and non-forest edge, yet 36.43% contributed to the TTE extent. We report the overall extent of the TTE (3.032 million km2) across North America and Greenland (53%), and Eurasia (47%), where 0.697 million km2 is non-forest edge, 0.549 million km2 is sparse forest, and 1.787 million km2 is open canopy forest. Diffuse forest landscapes dominate the TTE (79%), and abrupt landscapes (~19%) indicate portions of the TTE where sparse forest and non-forest edge are the prevailing structural patterns. This account of the TTE quantifies the area of the cold edge of the boreal forest where previous global estimates show high discrepancies, and can help target monitoring and prediction of circumpolar dynamics.

Paul M Montesano↗

Efficient numerical methods to solve sparse linear equations with application to PageRank

Over the last two decades, the PageRank problem has received increased interest from the academic community as an efficient tool to estimate web-page importance in information retrieval. Despite numerous developments, the design of efficient optimization algorithms for the PageRank problem is still a challenge. Here, we propose three new algorithms with a linear time complexity for solving the problem over a bounded-degree graph. The idea behind them is to set up the PageRank as a convex minimization problem over a unit simplex, and then solve it using iterative methods with small iteration complexity. Our theoretical results are supported by an extensive empirical justification using real-world and simulated data.

97 MATHEMATICS AND COMPUTING↗

Processing Aleatory and Epistemic Uncertainties in Experimental Data From Sparse Replicate Tests of Stochastic Systems for Real-Space Model Validation

This paper presents a practical methodology for propagating and processing uncertainties associated with random measurement and estimation errors (that vary from test-to-test) and systematic measurement and estimation errors (uncertain but similar from test-to-test) in inputs and outputs of replicate tests to characterize response variability of stochastically varying test units. Also treated are test condition control variability from test-to-test and sampling uncertainty due to limited numbers of replicate tests. These aleatory variabilities and epistemic uncertainties result in uncertainty on computed statistics of output response quantities. The methodology was developed in the context of processing experimental data for “real-space” (RS) model validation comparisons against model-predicted statistics and uncertainty thereof. The methodology is flexible and sufficient for many types of experimental and data uncertainty, offering the most extensive data uncertainty quantification (UQ) treatment of any model validation method the authors are aware of. It handles both interval and probabilistic uncertainty descriptions and can be performed with relatively little computational cost through use of simple and effective dimension- and order-adaptive polynomial response surfaces in a Monte Carlo (MC) uncertainty propagation approach. A key feature of the progressively upgraded response surfaces is that they enable estimation of propagation error contributed by the surrogate model. Sensitivity analysis of the relative contributions of the various uncertainty sources to the total uncertainty of statistical estimates is also presented. Finally, the methodologies are demonstrated on real experimental validation data involving all the mentioned sources and types of error and uncertainty in five replicate tests of pressure vessels heated and pressurized to failure. Simple spreadsheet procedures are used for all processing operations.

97 MATHEMATICS AND COMPUTING↗

The increase in animal mortality risk following exposure to sparsely ionizing radiation is not linear quadratic with dose

The US government regulates allowable radiation exposures relying, in large part, on the seventh report from the committee to estimate the Biological Effect of Ionizing Radiation (BEIR VII), which estimated that most contemporary exposures- protracted or low-dose, carry 1.5 fold less risk of carcinogenesis and mortality per Gy than acute exposures of atomic bomb survivors. This correction is known as the dose and dose rate effectiveness factor for the life span study of atomic bomb survivors (DDREF LSS ). As a result, it was calculated by applying a linear-quadratic dose response model to data from Japanese atomic bomb survivors and a limited number of animal studies.

63 RADIATION, THERMAL, AND OTHER ENVIRON. POLLUTAN↗

Charge exchange reactions and applications to astrophysics

Neutrino-induced reactions have been known to play important roles as the neutrino process on the nucleosynthesis in core collapsing supernovae (SNe) explosions because expected neutrino flux and energy are sufficiently high enough to excite many relevant nuclei in spite of small cross sections of the weak interaction. However, we do not have enough data for the neutrino reaction to be exploited in the network calculation. Only a sparse data in the relevant energy range is known, in specific, for {sup 12}C. Therefore we have to rely on theoretical estimation of the reaction, which has two different modes, charge current (CC) and neutral current (NC). In particular, CC reactions are closely related to charge exchange reactions (CEXRs) which are feasible in the experiment, such as, (p,n) or (n,p) reactions. These CEXRs are usually dominated by the Gamow-Teller (GT) transition in the lower energy region. In this respect, any theoretical approaches for the neutrino reaction should be investigated for the CEXR because we have and expect more useful experimental data. After confirming our models to the GT strength deduced from the CEXR, we calculated neutrino-induced reactions in the energy range below the quasielastic region for nuclei of astrophysical importance. Our calculations are carried out with the Quasi-particle Random Phase Approximation (QRPA), which successfully described the nuclear beta decays of relevant nuclei. To describe neutrino-nucleus reactions, general multipole transitions by the weak interaction are considered for CC and NC reactions. Both reactions are described in a theoretical framework. Our results are shown to well reproduce the data from CEXRs and the sparse experimental data related to the neutrino-induced reaction, and further extended for neutrino reactions on various nuclear targets. Parts of the results are reported in this talk.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

Derivative-based SINDy (DSINDy): Addressing the challenge of discovering governing equations from noisy data

Recent advances in the field of data-driven dynamics allow for the discovery of ODE systems using state measurements. One approach, known as Sparse Identification of Nonlinear Dynamics (SINDy), assumes the dynamics are sparse within a predetermined basis in the states and finds the expansion coefficients through linear regression with sparsity constraints. This approach requires an accurate estimation of the state time derivatives, which is not necessarily possible in the high-noise regime without additional constraints. We present an approach called Derivative-based SINDy (DSINDy) that combines two novel methods to improve ODE recovery at high-noise levels. First, we denoise the state variables by applying a projection operator that leverages the assumed basis for the system dynamics. Second, we use a second order cone program (SOCP) to find the derivative and governing equations simultaneously. We derive theoretical results for the projection-based denoising step, which allow us to estimate the values of hyperparameters used in the SOCP formulation. This underlying theory helps limit the number of required user-specified parameters. Finally, we present results demonstrating that our approach leads to improved system recovery for the Van der Pol oscillator, the Duffing oscillator, the Rössler attractor, and the Lorenz 96 model.

97 MATHEMATICS AND COMPUTING↗

Adaptive sparse polynomial chaos expansion based on least angle regression

Polynomial chaos (PC) expansions are used in stochastic finite element analysis to represent the random model response by a set of coefficients in a suitable (so-called polynomial chaos) basis. The number of terms to be computed grows dramatically with the size of the input random vector, which makes the computational cost of classical solution schemes (may it be intrusive (i.e. of Galerkin type) or non intrusive) unaffordable when the deterministic finite element model is expensive to evaluate. To address such problems, the paper describes a non intrusive method that builds a sparse PC expansion. First, an original strategy for truncating the PC expansions, based on hyperbolic index sets, is proposed. Then an adaptive algorithm based on least angle regression (LAR) is devised for automatically detecting the significant coefficients of the PC expansion. Beside the sparsity of the basis, the experimental design used at each step of the algorithm is systematically complemented in order to avoid the overfitting phenomenon. The accuracy of the PC metamodel is checked using an estimate inspired by statistical learning theory, namely the corrected leave-one-out error. As a consequence, a rather small number of PC terms are eventually retained (sparse representation), which may be obtained at a reduced computational cost compared to the classical 'full' PC approximation. The convergence of the algorithm is shown on an analytical function. Then the method is illustrated on three stochastic finite element problems. The first model features 10 input random variables, whereas the two others involve an input random field, which is discretized into 38 and 30 - 500 random variables, respectively.

97 MATHEMATICS AND COMPUTING↗

Closure of Earth’s Global Seasonal Cycle of Energy Storage

Abstract The global seasonal cycle of energy in Earth’s climate system is quantified using observations and reanalyses. After removing long-term trends, net energy entering and exiting the climate system at the top of the atmosphere (TOA) should agree with the sum of energy entering and exiting the ocean, atmosphere, land, and ice over the course of an average year. Achieving such a balanced budget with observations has been challenging. Disagreements have been attributed previously to sparse observations in the high-latitude oceans. However, limiting the local vertical integration of new global ocean heat content estimates to the depth to which seasonal heat energy is stored, rather than integrating to 2000 m everywhere as done previously, allows closure of the global seasonal energy budget within statistical uncertainties. The seasonal cycle of energy storage is largest in the ocean, peaking in April because ocean area is largest in the Southern Hemisphere and the ocean’s thermal inertia causes a lag with respect to the austral summer solstice. Seasonal cycles in energy storage in the atmosphere and land are smaller, but peak in July and September, respectively, because there is more land in the Northern Hemisphere, and the land has more thermal inertia than the atmosphere. Global seasonal energy storage by ice is small, so the atmosphere and land partially offset ocean energy storage in the global integral, with their sum matching time-integrated net global TOA energy fluxes over the seasonal cycle within uncertainties, and both peaking in April.

54 ENVIRONMENTAL SCIENCES↗

Assimilation of Combined Microwave and Lightning Measurement in a Mesoscale Weather Prediction Model

Intermittent measurements of precipitation and integrated water vapor (IWV) distributions were retrieved from the Special Sensor Microwave/Imager (SSM/I) and Tropical Rainfall Measuring Mission (TRMM) Microwave Imager (TMI) radiometers. Lightning generates very low frequency (VLF) radio noise pulses called sferics. Those pulses propagate over large distances so that they can be continuously monitored with a sparse network of ground based radio receivers. Sferics data, tuned with intermittent spaceborne microwave radiometer data, were used to generate estimated rainfall that was assimilated into a mesoscale weather prediction model. Both continuous latent heating adjustment and a variational technique are applied as assimilation procedures to evaluate the impact of lightning observations on the forecast of an intense winter squall line over the Gulf of Mexico. Sensitivities to the assimilation of additional measurements such as IWV and sea surface temperature (SST), and measurement errors will also be discussed.

Chang, Dong-Eon↗

Bayesian Geostatistical Modelling of PM10 and PM2.5 Surface Level Concentrations in Europe Using High-Resolution Satellite-Derived Products

Air quality monitoring across Europe is mainly based on in situ ground stations which are too sparse to accurately assess the exposure effects of air pollution for the entire continent. The demand for precise predictive modelsthat estimate gridded geophysical parameters of ambient air at high spatial resolution has rapidly grown. Here, we investigate the potential of satellite derived products to improve particulate matter (PM) estimates. Bayesiangeostatistical models addressing confounding between the spatial distribution of pollutants and remotely sensed predictors were developed to estimate yearly averages of both, fine (PM2.5) and coarse (PM10) surface PM concentrations at 1 sq.km spatial resolution over 46 European countries and were compared to geostatistical, geographically weighted and land-use regression formulations. Rigorous model selection identified the Earth observation data which contribute most to pollutants' estimation. Geostatistical models outperformed the predictive ability of the frequently employed land-use regression. The resulting estimates of PM10 and PM2.5, which represent the main air quality indicators for the urban Sustainable Development Goal, indicate that in 2016, 66.2% of the European population was breathing air above the WHO Air Quality Guidelines thresholds. Our estimates are readily available to policy makers and scientists assessing the effects of long-term exposure to pollution on human and ecosystem health.

Beloconi, Anton↗

Inferring the thermal-infrared hemispheric emission from a sparsely-vegetated surface by directional measurements

The thermal-infrared (longwave) emission from a vegetated terrain is generally anisotropic, i.e., the emission temperature varies with the view direction. If a directional measurement of temperature is considered to be equal to the effective temperature of the hemispheric emission, then the estimate of the latter can be significantly in error. The view-direction (zenith angle theta(sub eq) at which the emission equivalence does hold is determined in our modeling study. In a two-temperature field-of-view (soil and plants), theta(sub eq) falls in a narrow range depending on plant density and canopy architecture. Theta(sub eq) does not depend on soil and (uniform) plant temperatures nor on their ratio, even though the pattern of emission vs. the view direction depends crucially on this ratio. For a sparse canopy represented as thin, vertical cylindrical stalks (or vertical blades uniformly distributed in azimuth) with horizontal facets, theta(sub eq) ranges from 48 to 53 deg depending on the optical density of the vertical elements alone. When plant elements are modeled as small spheres, theta(sub eq) lies between 53 to 57 deg (for the same values of the canopy optical density). Only for horizontal leaves (a truly planophile canopy) is the temperature measured from any direction equal to the temperature of the hemispheric emission. When the emission temperature changes with optical depth within the canopy at a specified rate, theta(sub eq) depends to some extent on that rate. For practically any sparsely vegetated surface, a directional measurement at the zenith angle of 50 deg offers an appropriate evaluation of the hemispheric emission, since the error in the estimate will, at most, only slightly exceed 1% (around 4 W/sq m). Estimates of the hemispheric emission through a nadir measurement, on the other hand, can be in error in some cases by about 10%, i.e., on the order of 40 W/sq m.

Otterman, J.↗

Architecture Independent Integrated Early Performance and Energy Estimation

The end of Dennard’s scaling, coupled to the necessity to keep increasing computational performance in constrained power envelopes, is leading developers towards the need to opti- mize applications not only for performance, but also for energy consumption. In such a scenario, solutions that enable estimating performance and energy consumption of a program as early as possible, even during development and without the necessity to access the final target hardware, can become fundamental tools to reduce the design space, the optimization effort, and provide potential opportunities to automate the optimization process itself. In this position paper, we discuss the design of a performance and energy estimation framework based on a retargetable com- piler. Our proposed approach targets the Intermediate Repre- sentation (IR) of the LLVM compiler. The IR based approach enables, after a model is built, estimation from a host different from the target architecture, even considering dynamic infor- mation. We present the rationale behind such a framework, identify opportunities and components readily available (such as the provided static analysis, quick ways to add instrumentation for dynamic profiling, and the multiple backends), and aspects that still need further research efforts (such as more effective non-linear, or machine learning based models). We also discuss a hand-developed case study, based on a simple sparse matrix vector multiplication with a linear model, to motivate the needs for such a framework.

Estimation, Compilers, Performance Analysis↗

Context-aware learning of hierarchies of low-fidelity models for multi-fidelity uncertainty quantification

Multi-fidelity Monte Carlo methods leverage low-fidelity and surrogate models for variance reduction to make tractable uncertainty quantification even when numerically simulating the physical systems of interest with high-fidelity models is computationally expensive. This work proposes a context-aware multi-fidelity Monte Carlo method that optimally balances the costs of training low-fidelity models with the costs of Monte Carlo sampling. It generalizes the previously developed context-aware bi-fidelity Monte Carlo method to hierarchies of multiple models and to more general types of low-fidelity models. When training low-fidelity models, the proposed approach takes into account the context in which the learned low-fidelity models will be used, namely for variance reduction in Monte Carlo estimation, which allows it to find optimal trade-offs between training and sampling to minimize upper bounds of the mean-squared errors of the estimators for given computational budgets. This is in stark contrast to traditional surrogate modeling and model reduction techniques that construct low-fidelity models with the primary goal of approximating well the high-fidelity model outputs and typically ignore the context in which the learned models will be used in upstream tasks. Further, the proposed context-aware multi-fidelity Monte Carlo method applies to hierarchies of a wide range of types of low-fidelity models such as sparse-grid and deep-network models. Numerical experiments with the gyrokinetic simulation code Gene show speedups of up to two orders of magnitude compared to standard estimators when quantifying uncertainties in small-scale fluctuations in confined plasma in fusion reactors. This corresponds to a runtime reduction from 72 days to four hours on one node of the Lonestar6 supercomputer at the Texas Advanced Computing Center.

42 ENGINEERING↗

Recommendations for Using Noise Monitors to Estimate Noise Exposure During X-59 Community Tests

A low fidelity simulation approach is used to explore how to place and use noise monitors during X-59 QueSST community tests, where people’s annoyance to the noise produced by the X-59 aircraft will be gathered. Several recommendations are provided including: 1) the desired number of sparsely spaced noise monitor sites within the survey area, 2) whether to group and average measurements across multiple noise monitors located at a site, 3) what spacing should be used if grouped noise monitors are used, 4) an approach to mitigate ambient noise contamination at the measurement sites, 5) a method to combine empirical and predicted dose estimates to provide a single dose estimate for respondents, and 6) assessing how changes in turbulence intensity and array configuration affect dose uncertainty. To make these recommendations, the error that is expected when fitting contrived, smoothly varying sonic boom “reference exposure surfaces” is studied when a spatially sparse and scattered set of samples is used as responses for the fit. The reference exposure surfaces mimic the sonic boom exposure at ground level that might be expected in the X-59 survey area in the absence of atmospheric turbulence, ambient noise, and other localized effects. The spatial extent of these surfaces varies and is representative of the different survey area sizes that might be expected during future X-59 community overflight tests. These contrived reference surfaces are sampled, and those reference samples are then perturbed to mimic atmospheric turbulence, ambient noise and other localized effects that might affect noise monitor measurements within overflown communities. Two different surface fitting methods are investigated when fitting these perturbed samples to approximate the reference surface. The first method uses interpolation between the perturbed data at the scattered sites to compute the fit. The second method fits a polynomial surface model to the perturbed data using ordinary least squares regression analysis. For both fitting methods, the root mean square fit error is computed from the pointwise difference between the fit surface and the reference surface as the count and configuration of the sites is varied while also averaging the error across many different realizations of both the smooth variation of the reference exposure surface and the random, localized perturbations at the sample sites. Different site configurations are compared using this error statistic to make the recommendations noted above. Additionally, the two fitting approaches (interpolation vs linear regression) are compared based on the fit error observed in these simulations. These analyses, comparisons, and recommendations should inform future decisions on the noise monitor placement and the methods used to analyze the noise monitor data that is collected during X-59 community overflights.

sonic boom↗

Uranium-series comminution ages of continental sediments: Case study of a Pleistocene alluvial fan

Obtaining quantitative information about the timescales associated with sediment transport, storage, and deposition in continental settings is important but challenging. The uranium-series comminution age method potentially provides a universal approach for direct dating of Quaternary detrital sediments, and can also provide estimates of the sediment transport and storage timescales. (The word"comminution" means"to reduce to powder," reflecting the start of the comminution age clock as reduction of lithic parent material below a critical grain size threshold of ~;;50 mu m.) To test the comminution age method as a means to date continental sediments, we applied the method to drill-core samples of the glacially-derived Kings River Fan alluvial deposits in central California. Sediments from the 45 m core have independently-estimated depositional ages of up to ~;;800 ka, based on paleomagnetism and correlations to nearby dated sediments. We characterized sequentially-leached core samples (both bulk sediment and grain size separates) for U, Nd, and Sr isotopes, grain size, surface texture, and mineralogy. In accordance with the comminution age model, where 234U is partially lost from small sediment grains due to alpha recoil, we found that (234U/238U) activity ratios generally decrease with age, depth, and specific surface area, with depletions of up to 9percent relative to radioactive equilibrium. The resulting calculated comminution ages are reasonable, although they do not exactly match age estimates from previous studies and also depend on assumptions about 234U loss rates. The results indicate that the method may be a significant addition to the sparse set of available tools for dating detrital continental sediments, following further refinement. Improving the accuracy of the method requires more advanced models or measurements for both the recoil loss factor fa and weathering effects. We discuss several independent methods for obtaining fa on individual samples that may be useful for future studies.

58 GEOSCIENCES↗

Herbarium records provide reliable phenology estimates in the understudied tropics

1. Plant phenology has been shifting dramatically in response to climate change, a shift that may have significant and widespread ecological consequences. Of particular concern are tropical biomes, which represent the most biodiverse and imperilled regions of the world. However, compared to temperate floras, we know little about phenological responses of tropical plants because long-term observational datasets from the tropics are sparse. 2. Herbarium specimens have greatly increased our phenological knowledge in temperate regions, but similar data have been underutilized in the tropics and their suitability for this purpose has not been broadly validated. Here, we compare phenological estimates derived from field observational data (i.e. plot surveys) and herbarium specimens at various spatial and taxonomic scales to determine whether specimens can provide accurate estimations of reproductive timing and its spatial variation. 3. Here, we demonstrate that phenological estimates from field observations and herbarium specimens coincide well. Fewer than 5% of the species exhibited significant differences between flowering periods inferred from field observations versus specimens regardless of spatial aggregation. In contrast to studies based on field records, herbarium specimens sampled much larger geographic and climatic ranges, as has been documented previously for temperate plants, and effectively captured phenological responses across varied environments. 4. Synthesis. Herbarium specimens are verified to be a vital resource for closing the gap in our phenological knowledge of tropical systems. Tropical plant reproductive phenology inferred from herbarium records is widely congruent with field observations, suggesting that they can and should be used to investigate phenological variation and their associated environmental cues more broadly across tropical biomes.

Brazil↗