Composite Time Series Research Articles

Evaluating the Impact of Abrupt Changes in Forest Policy and Management Practices on Landscape Dynamics: Analysis of a Landsat Image Time Series in the Atlantic Northern Forest

Sustainable forest management is based on functional relationships between management actions, landscape conditions, and forest values. Changes in management practices make it fundamentally more difficult to study these relationships because the impacts of current practices are difficult to disentangle from the persistent influences of past practices. Within the Atlantic Northern Forest of Maine, U.S.A., forest policy and management practices changed abruptly in the early 1990s. During the 1970s-1980s, a severe insect outbreak stimulated salvage clearcutting of large contiguous tracts of spruce-fir forest. Following clearcut regulation in 1991, management practices shifted abruptly to near complete dependence on partial harvesting. Using a time series of Landsat satellite imagery (1973-2010) we assessed cumulative landscape change caused by these very different management regimes. We modeled predominant temporal patterns of harvesting and segmented a large study area into groups of landscape units with similar harvest histories. Time series of landscape composition and configuration metrics averaged within groups revealed differences in landscape dynamics caused by differences in management history. In some groups (24% of landscape units), salvage caused rapid loss and subdivision of intact mature forest. Persistent landscape change was created by large salvage clearcuts (often averaging > 100 ha) and conversion of spruce-fir to deciduous and mixed forest. In groups that were little affected by salvage (56% of landscape units), contemporary partial harvesting caused loss and subdivision of intact mature forest at even greater rates. Patch shape complexity and edge density reached high levels even where cumulative harvest area was relatively low. Contemporary practices introduced more numerous and much smaller patches of stand-replacing disturbance (typically averaging <15 ha) and a correspondingly large amount of edge. Management regimes impacted different areas to different degrees, producing different trajectories of landscape change that should be recognized when studying the impact of policy and management practices on forest ecology.

Increasing the quality, comparability and accessibility of phytoplankton species composition time-series data

Phytoplankton diversity and its variation over an extended time scale can provide answers to a wide range of questions relevant to societal needs. These include human health, the safe and sustained use of marine resources and the ecological status of the marine environment, including long-term changes under the impact of multiple stressors. The analysis of phytoplankton data collected at the same place over time, as well as the comparison among different sampling sites, provide key information for assessing environmental change, and evaluating new actions that must be made to reduce human induced pressures on the environment. To achieve these aims, phytoplankton data may be used several decades later by users that have not participated in their production, including automatic data retrieval and analysis. The methods used in phytoplankton species analysis vary widely among research and monitoring groups, while quality control procedures have not been implemented in most cases. Here we highlight some of the main differences in the sampling and analytical procedures applied to phytoplankton analysis and identify critical steps that are required to improve the quality and inter-comparability of data obtained at different sites and/or times. Harmonization of methods may not be a realistic goal, considering the wide range of purposes of phytoplankton time-series data collection. However, we propose that more consistent and detailed metadata and complementary information be recorded and made available along with phytoplankton time-series datasets, including description of the procedures and elements allowing for a quality control of the data. To keep up with the progress in taxonomic research, there is a need for continued training of taxonomists, and for supporting and complementing existing web resources, in order to allow a constant upgrade of knowledge in phytoplankton classification and identification. Efforts towards the improvement of metadata recording, data annotation and quality control procedures will ensure the internal consistency of phytoplankton time series and facilitate their comparability and accessibility, thus strongly increasing the value of the precious information they provide. Ultimately, the sharing of quality controlled data will allow one to recoup the high cost of obtaining the data through the multiple use of the time-series data in various projects over many decades.

Satellite view of quasi‐equilibrium states in tropical convection and precipitation microphysics

AbstractThis study shows the time series of statistical composites of precipitation‐microphysics signals derived from long‐term Tropical Rainfall Measuring Mission (TRMM) satellite observations aggregated over the entire tropical domain (37°N–37°S). The result shows the nearly time‐invariant monthly signal statistics throughout this time, confirming convection quasi‐equilibrium (CQE) states. Merged precipitation data, with much better temporal and spatial coverage, provide evidence that the equilibrium state occurs on a daily scale. These results further indicate the presence of precipitation microphysics quasi‐equilibrium (MQE) within the CQE environment. A simple analytic microphysics framework illustrates the equilibrium precipitation size distribution, as compared with the TRMM radar‐based as well as preliminary Global Precipitation Measurement combined retrievals. The MQE readily explains the near‐constant tropical precipitation rate, which is roughly balanced with atmospheric radiative cooling rate at the entire tropical scale. Further investigation is required through theoretical, observational, and numerical manners to support the MQE hypothesis.

Modeling Compositional Time Series with Vector Autoregressive Models

Multivariate time series describing relative contributions to a total (like proportional data) are called compositional time series. They need to be transformed first to the usual Euclidean geometry before a time series model is fitted. It is shown how an appropriate transformation can be chosen, resulting in coordinates with respect to the Aitchison geometry of compositional data. Using vector autoregressive models, the standard approach based on raw data is compared with the compositional approach based on transformed data. The results from the compositional approach are consistent with the relative nature of the observations, while the analysis of the raw data leads to several inconsistencies and artifacts. The compositional approach is extended to the case when also the total of the compositional parts is of interest. Moreover, a concise methodology for an interpretation of the coordinates in the transformed space together with the corresponding statistical inference (like hypotheses testing) is provided. Copyright © 2015 John Wiley &amp; Sons, Ltd.

Long-term variability of storm surge frequency in the Venice Lagoon: an update thanks to 18th century sea level observations

Abstract. Sea level observations made in the Venice Lagoon between 1751 and 1792 have been recovered, consisting of two time series of daily data on high and low waters in Venice and Chioggia. From comparisons with modern observations, the quality of the 18th century data appears to be good enough to allow a useful analysis. A composite time series of daily mean sea level is obtained by merging the 18th century data and 1872–2004 observations in Venice Punta della Salute. The absence of reliable information on vertical references prevents the connection of the two 18th century time series with each other and with modern observations. However, daily sea level anomalies relative to the mean sea level enable us to recognize storm surge events that appear to occur more frequently in the second half of the 18th century than in the late 19th and 20th centuries, particularly during the 1751–1769 period. The record-breaking storm surge of 4 November 1966 turns out to be a remarkable event even in comparison to the events extracted from the 18th century time series. Further work is required to fill the gap between the old and modern observations.

Change Detection of Rice Cultivation in Bangladesh Based on the Phenological Analysis of MODIS Data

Mapping rice cultivation is indispensable for monitoring food supply conditions in Bangladesh because of the economical importance of the crop for supporting ever increasing population in the country. In this paper, we extract the rice paddy field using the MODIS satellite data for five districts of Pabna, Manikganj, Sherpur, Sylhet, and Gazipur, each of which is characterized with its own aspects in terms of rice cultivation. Land classification is implemented using the vegetation index information derived from the red (band 1) and near-infrared (band 2) bands of MODIS 8-day composite time series data for the two time periods of 2001-2003 and 2011-2013. Results of unsupervised classification indicate that the paddy area coverage increased about 4% and 1% in Gazipur and Sylhet, respectively. In Pabna, Manikganj, and Sherpur, on the other hand, paddy area decreased by 10%, 2% and 5%, respectively, whereas notable increase of 12%, 2% and 7% was found in homestead area coverage, which is becoming more and more important for better management of small-scale agroforestry. At the same time, in Sherpur and Sylhet, forest area increased by 1% and 2% over the same time period. As a validation of these results, the changes detected in Gazipur are compared with those previously derived from the analysis of Landsat data with higher spatial resolution of 30 m as compared with that of MODIS (250 m). Also, the seasonal rice cropping pattern is studied in these five districts for discriminating cultivated rice types. These changes suggest that as a whole, efforts are being made to increase the food production, though the influence of population pressure and economic growth is apparent in these regions.

A Spatio-temporal disaggregation method to derive time series of Normalized Difference Vegetation Index and Land Surface Temperature at fine spatial resolution

Abstract. Estimation of evapotranspiration (ET) from remote sensing based energy balance models have evolved as a promising tool in the field of water resources management. Performance of energy balance models and reliability of ET estimates is decided by the availability of remote sensing data at high spatial and temporal resolutions. However huge tradeoff in the spatial and temporal resolution of satellite images act as major constraints in deriving ET at fine spatial and temporal resolution using remote sensing based energy balance models. Hence a need exists to derive finer resolution data from the available coarse resolution imagery, which could be applied to deliver ET estimates at scales to the range of individual fields. The current study employed a spatio-temporal disaggregation method to derive fine spatial resolution (60 m) images of NDVI by integrating the information in terms of crop phenology derived from time series of MODIS NDVI composites with fine resolution NDVI derived from a single AWiFS data acquired during the season. The disaggregated images of NDVI at fine resolution were used to disaggregate MODIS LST data at 960 m resolution to the scale of Landsat LST data at 60 m resolution. The robustness of the algorithm was verified by comparison of the disaggregated NDVI and LST with concurrent NDVI and LST images derived from Landsat ETM+. The results showed that disaggregated NDVI and LST images compared well with the concurrent NDVI and LST derived from ETM+ at fine resolution with a high Nash Sutcliffe Efficiency and low Root Mean Square Error. The proposed disaggregation method proves promising in generating time series of ET at fine resolution for effective water management.

Constraints on the role of tectonic and climate on erosion revealed by two time series analysis of marine cores around New Zealand

Physical and chemical weathering govern rocks and chemical elements cycles at the Earth surface. The importance of climate and tectonic forcings has been identified but their relative influence still needs to be investigated. We address this question by studying neodymium (Nd) isotopic composition time series in two ocean sediment cores located on the eastern and western sides of New Zealand. We measured both detrital and authigenic phases of the sediment to decouple the oceanic from the continental contributions. The current and past climatic and oceanographic settings of this region have already been well studied, providing the underlying framework.The results show glacial–interglacial variations of the authigenic signal, with larger amplitudes for the eastern core. We argue that oceanic influences can be discarded and propose that continental erosion on the South Island controls the Nd isotopes records in both cores. Mixing calculations show that Nd discharge was 2 to 10 times higher during glacial times than during interglacials on the eastern side, whereas these variations are almost negligible on the western side. We suggest that, on the East, larger ice volumes and fluxes during glacials generate higher physical erosion and thus an increase of easily weathered fine grains. In contrast, we observe steady rates on the western side. This implies that where both rock uplift and precipitation are extreme, topography and thus erosion rates are less sensitive to switches from glacial to fluvial erosion. Furthermore, the temporal relationship between climatic indicators like oxygen isotopes and the Nd isotopic composition reveals a nonlinear response of erosion to climate forcing. Those results contrast with previous studies using the same tracer that indicated stronger erosion over the Himalaya during interglacial periods and point out the complexity of erosion and weathering responses to climatic and tectonic forcings.

Temporal trends in δ18O composition of precipitation in Germany: insights from time series modelling and trend analysis

AbstractTemporal and spatial variations of stable oxygen (18O) and hydrogen (2H) isotope measurements in precipitation act as important proxies for changing hydro‐meteorological and regional and global climate patterns. Temporal trends in time series of the stable isotope composition in precipitation were rarely observed, and they are poorly understood. These might be a result of a lack of proper trend detection tools and effort for exploring trend processes. Here, we investigate temporal trends of δ18O in precipitation at 17 observation stations in Germany between 1978 and 2009. We test if significant trends in the isotope time series from different models can be observed. Mann–Kendall trend tests are applied on the isotope series, using general multiplicative seasonal autoregressive integrate moving average (ARIMA) models, which account for first and higher order serial correlations. Effects of temperature, precipitation, and geographic parameters on isotope trends are also investigated in the proposed models. To benchmark our proposed approach, the ARIMA results are compared with a trend‐free pre‐whitening procedure, the state of the art method for removing the first order autocorrelation in environmental trend studies. Moreover, we further explore whether higher order serial correlations in isotope series affects our trend results. Overall, three out of the 17 stations show significant changes when higher order autocorrelation are adjusted, and four show a significant trend when temperature and precipitation effects are considered. The significant trends in the isotope time series generally occur only at low elevation stations. Higher order autoregressive processes are shown to be important in the isotope time series analysis. Results suggest that the widely used trend analysis with only the first order autocorrelation adjustment may not adequately take account of the high order autocorrelated processes in the stable isotope series. The investigated time series analysis method including higher autocorrelation and external climate variable adjustments is shown to be a better alternative. Copyright © 2014 John Wiley &amp; Sons, Ltd.

Evolution of Mediterranean sea surface temperatures 3.5–1.5 Ma: Regional and hemispheric influences

We present a composite time series of Mediterranean sea surface temperature (SST) and marine biomarker accumulation for the time span from 3.5 to 1.525 Ma, based on alkenone unsaturation and concentration from hemipelagic sediments outcropping in southern Italy and Sicily. Paleotemperature data define three regimes: a late Pliocene climate on average 4–5°C warmer than modern, a latest Pliocene to early Pleistocene onset of 41 kyr cycles, and a major increase in the range of glacial–interglacial temperature change at ∼1.84 Ma that shortly precedes the former definition of the Plio-Pleistocene boundary in the Crotone sequence. Pliocene sea surface temperature (SST) cycles are dominated by precession, with a ∼1.5°C range. Obliquity-related rhythms influence SST significantly shortly after ∼2.8 Ma (equivalent to MIS G10) and dominate after ∼2.51 Ma (equivalent to MIS 100). However, little, if any, long-term cooling occurred on an interglacial basis until after ∼1.85 Ma. Alkenone concentrations provide a good proxy for the accumulation of marine organic matter, and primarily reflect regional hydrology. Organic sedimentation, including the formation of layers highly enriched in organic matter (“sapropels”) was paced throughout by precessional variations despite changes in both average regional temperature, and the shift in temperature variance to the 41 kyr obliquity cycle in the latest Pliocene and early Pleistocene. Our reconstruction therefore highlights the intermingling of both hemispheric-wide changes in temperature and regional variations in the hydrological cycle that combined to force major evolutionary changes in the fauna and flora of northern Africa and the southern Mediterranean in late Pliocene to mid-Pleistocene time.

Preliminary analysis of acceleration of sea level rise through the twentieth century using extended tide gauge data sets (August 2014)

This work explores the potential for extending tide gauge time series from the Permanent Service for Mean Sea Level (PSMSL) using historical documents, PSMSL ancillary data, and by developing additional composite time series using near neighbor tide gauges. The aim was to increase the number, completeness, and geographical extent of records covering most or all of the twentieth century. The number of at least 75% complete century-scale time series have been approximately doubled over the original PSMSL data set. In total, over 4800 station years have been added, with 294 of these added to 10 long Southern Hemisphere records. Individual century-scale acceleration values derived from this new extended data set tend to converge on a value of 0.01 ± 0.008 mm/yr2. This result agrees closely with recent work and is statistically significant at the 1 sigma level. Possible causes of acceleration and errors are briefly discussed. Results confirm the importance of current data archeology projects involving digitization of the remaining archives of hard copy tide gauge data for sea level and climate studies.

CO2 Adsorption: Experimental Investigation and CFD Reactor Model Validation

The National Energy Technology Laboratory is investigating a new process for CO2 capture from large sources such as utility power generation facilities as an alternative to liquid amine based adsorption processes. Many of these advanced dry processes are based upon sorbents composed of supported polyamines. In this analysis, experiments have been conducted in a laboratory-scale fluidized bed reactor and compared to CFD reactor predictions using kinetics obtained from TGA tests. Batch experiments were conducted by flowing a mixture of CO2, H2O, and N2 (simulated flue gas) through a fluidized bed of sorbent material. The exit gas composition time series data is compared to CFD simulations using a 3-dimensional nonisothermal reacting multiphase flow model. The effects of the gas flow rate, distributor design, and particle size are explored through the CFD simulations. It is shown that the time duration for CO2 adsorption decreased for an increase in the gas flow. Fluid bed hydrodynamics indicated that there were regions in the reactor where the inert FCC particles segregated and defluidized; without adversely affecting the capacity of the sorbent to adsorb CO2. The details of the experimental facility and the model as well as the comparative analysis between the data and the simulation results are discussed.

Monitoring of Vegetation Dynamics in the Mongolia Using MODIS NDVIs and their Relationship to Rainfall by Natural Zone

Vegetation plays a very important role conserving the natural environment and improving the living environment for humans. Thus, investigating spatio-temporal changes in vegetation is a crucial indicator to understand natural environmental changes. We explored the spatio-temporal trends of changing vegetation cover in Mongolia from 2002 to 2010 by investigating changes in the normalized difference vegetation index (NDVI) with rainfall. The 16-day composite NDVI time series Terra MODIS (MOD13A2) and the gridded rainfall data of the Inter-governmental Panel on Climate Change (IPCC) were used to identify the relationship between vegetation dynamics and rainfall. Residual NDVI trends were computed. Following our general procedure, monthly NDVI and annual rainfall were correlated for each pixel. The resulting regression equation represented the statistical association between observed NDVI and rainfall, and the model predicted NDVI according to rainfall. Residuals of NDVI (differences between the observed and predicted NDVI) for each pixel were calculated, and the trends in these residuals was analyzed by linear regression. From the 12 months NDVI and rainfall values they determine a linear regression line for each pixel. The positive or negative slope of this line is considered to reflect an increase or decrease in green biomass. Our results show that vegetation cover changes in Mongolia were caused by both natural factors and human activities. The natural vegetation change, such as those in forests was influenced by climate change, whereas human activities were the main reason for the change in vegetation planted.

A Mechanism of Tropical Convection Inferred from Observed Variability in the Moist Static Energy Budget

Abstract Temporal variability in the moist static energy (MSE) budget is studied with measurements from a combination of different satellites including the Tropical Rainfall Measuring Mission (TRMM) and A-Train platforms. A composite time series before and after the development of moist convection is obtained from the observations to delineate the evolution of MSE and moisture convergences and, in their combination, gross moist stability (GMS). A new algorithm is then applied to estimate large-scale vertical motion from energy budget constraints through vertical-mode decomposition into first and second baroclinic modes and a background shallow mode. The findings are indicative of a possible mechanism of tropical convection. A gradual destabilization is brought about by the MSE convergence intrinsic to the positive second baroclinic mode (congestus mode) that increasingly counteracts a weak MSE divergence in the background state. GMS is driven to nearly zero as the first baroclinic mode begins to intensify, accelerating the growth of vigorous large-scale updrafts and deep convection. As the convective burst peaks, the positive second mode switches to the negative mode (stratiform mode) and introduces an abrupt rise in MSE divergence that likely discourages further maintenance of deep convection. The first mode quickly dissipates and GMS increases away from zero, eventually returning to the background shallow-mode state. A notable caveat to this scenario is that GMS serves as a more reliable metric when defined with a radiative heating rate included to offset MSE convergence.

Characteristics of the transmission of autoregressive sub-patterns in financial time series.

There are many types of autoregressive patterns in financial time series, and they form a transmission process. Here, we define autoregressive patterns quantitatively through an econometrical regression model. We present a computational algorithm that sets the autoregressive patterns as nodes and transmissions between patterns as edges, and then converts the transmission process of autoregressive patterns in a time series into a network. We utilised daily Shanghai (securities) composite index time series to study the transmission characteristics of autoregressive patterns. We found statistically significant evidence that the financial market is not random and that there are similar characteristics between parts and whole time series. A few types of autoregressive sub-patterns and transmission patterns drive the oscillations of the financial market. A clustering effect on fluctuations appears in the transmission process, and certain non-major autoregressive sub-patterns have high media capabilities in the financial time series. Different stock indexes exhibit similar characteristics in the transmission of fluctuation information. This work not only proposes a distinctive perspective for analysing financial time series but also provides important information for investors.

Effects of super typhoons on cyclonic ocean eddies in the western North Pacific: A satellite data‐based evaluation between 2000 and 2008

AbstractA composite time series of the merged satellite altimeters sea surface height anomaly (SSHA) data and satellite‐observed sea surface temperature (SST) data were used to identify eddies in the Western North Pacific Ocean (WNPO), where there were numbers of intense typhoons. This study systematically investigated 15 super typhoons during the period of 2000‐2008 in the WNPO to study their impacts on the pre‐typhoon ocean features, e.g., the cyclonic ocean eddy (COE) feature (closed contours of SSHA &lt; −6 cm) and neutral condition (SSHA between −6 and 6 cm). Two new COEs are generated by two super typhoons, and 18 pre‐existing COEs are intensified by 13 super typhoons. 5 of the 13 super typhoons each influenced two pre‐exisiting COEs. Although the typhoon‐induced maximum cooling centers had a right bias along the tracks due to wind conditions, pre‐existing COEs also play a significant role in determining the strength and location of large SST cooling. Three possible factors (maximum wind speed, typhoon translation speed and the typhoon forcing time, Tf) are employed to explain the interactions. Above all, the changes of the COE geometric and physical parameters (e.g., effective radius, area, SST, SSHA, and eddy kinetic energy) were mostly related to the typhoon forcing time, Tf. This is because Tf is a parameter that is a combination of the typhoon's translation speed, intensity and size. Although the typhoons may significantly impact COEs, such samples were not commonly observed. Thus, the impact of typhoon on the strength of COEs is generally inefficient.

An Independent Review of Existing Total Solar Irradiance Records

With the advent of space-based total solar irradiance (TSI) observations about 35 years ago, researchers’ understanding of solar variability and its causes has greatly improved. Controversies regarding the cross-calibration of the data from various TSI instruments have resulted in many different TSI composite time series. These composites agree well with each other on timescales ranging from days to years, but due to the limited stability of the instruments contributing to the composites, their quality is not yet sufficient to unambiguously detect possible changes between subsequent cycle minima. In this paper, the construction of the three most prominent TSI composite time series and the underlying TSI models is addressed. The difficulties associated with the cross-calibration of the data are considered, and the viewpoints of the different groups involved in the development of the composites are discussed.

Spatial-temporal survey and occupancy-abundance modeling to predict bacterial community dynamics in the drinking water microbiome.

ABSTRACTBacterial communities migrate continuously from the drinking water treatment plant through the drinking water distribution system and into our built environment. Understanding bacterial dynamics in the distribution system is critical to ensuring that safe drinking water is being supplied to customers. We present a 15-month survey of bacterial community dynamics in the drinking water system of Ann Arbor, MI. By sampling the water leaving the treatment plant and at nine points in the distribution system, we show that the bacterial community spatial dynamics of distance decay and dispersivity conform to the layout of the drinking water distribution system. However, the patterns in spatial dynamics were weaker than those for the temporal trends, which exhibited seasonal cycling correlating with temperature and source water use patterns and also demonstrated reproducibility on an annual time scale. The temporal trends were driven by two seasonal bacterial clusters consisting of multiple taxa with different networks of association within the larger drinking water bacterial community. Finally, we show that the Ann Arbor data set robustly conforms to previously described interspecific occupancy abundance models that link the relative abundance of a taxon to the frequency of its detection. Relying on these insights, we propose a predictive framework for microbial management in drinking water systems. Further, we recommend that long-term microbial observatories that collect high-resolution, spatially distributed, multiyear time series of community composition and environmental variables be established to enable the development and testing of the predictive framework.

Acidity decline in Antarctic ice cores during the Little Ice Age linked to changes in atmospheric nitrate and sea salt concentrations

AbstractAcidity is an important chemical variable that impacts atmospheric and snowpack chemistry. Here we describe composite time series and the spatial pattern of acidity concentration (Acy = H+ − HCO3−) during the last 2000 years across the Dronning Maud Land region of the East Antarctic Plateau using measurements in seven ice cores. Coregistered measurements of the major ion species show that sulfuric acid (H2SO4), nitric acid (HNO3), and hydrochloric acid (HCl) determine greater than 98% of the acidity value. The latter, also described as excess chloride (ExCl−), is shown mostly to be derived from postdepositional diffusion of chloride with little net gain or loss from the snowpack. A strong inverse linear relationship between nitrate concentration and inverse accumulation rate provides evidence of spatially homogenous fresh snow concentrations and reemission rates of nitrate from the snowpack across the study area. A decline in acidity during the Little Ice Age (LIA, 1500–1900 Common Era) is observed and is linked to declines in HNO3 and ExCl− during that time. The nitrate decline is found to correlate well with published methane isotope data from Antarctica (δ13CH4), indicating that it is caused by a decline in biomass burning. The decrease in ExCl− concentration during the LIA is well correlated to published sea surface temperature reconstructions in the Atlantic Ocean, which suggests increased sea salt aerosol production associated with greater sea ice extent.

Dendroclimatic reconstructions from multiple co‐occurring species: a case study from an old‐growth deciduous forest in Indiana, USA

ABSTRACTTree‐ring‐based climate reconstructions are typically derived from either (1) a single species from one or multiple locations or (2) multiple species from multiple locations. Here, we investigate the ability of using multiple co‐occurring canopy‐dominant species from a single location for climate reconstructions based in the eastern United States. Using a variety of techniques, we first compare the climate signals of three canopy‐dominant species (Quercus rubra, Quercus alba, Liriodendron tulipifera) at an old‐growth forest in southern Indiana. We then determine if a composite time series of these co‐occurring species increases or decreases the reconstruction model skill. Climate–growth correlation analysis of the species reveals strong and consistent relationships with summer [June–August (JJA)] Palmer Drought Severity Index (PDSI) during the period 1895–2012. We first use a split‐sample reconstruction technique to compare the performance of species reconstruction models. We then use a nested technique to build a composite chronology against which to compare the individual species chronologies. The composite chronology of all three species accounts for up to 38% of the mean variation in JJA PDSI, and verification statistics indicate robust statistical skill from 1870 to 2012. The composite chronology reconstruction also outperforms each individual species model, indicating that using multiple co‐occurring species increases reconstruction skill, at least from a single study site. Furthermore, model performance is improved by using nested reconstruction techniques, and implicates the potential ability to use multiple co‐occurring species across multiple locations in the eastern United States.