Climate Network Research Articles

Causal Discovery for Climate Research Using Graphical Models

AbstractCausal discovery seeks to recover cause–effect relationships from statistical data using graphical models. One goal of this paper is to provide an accessible introduction to causal discovery methods for climate scientists, with a focus on constraint-based structure learning. Second, in a detailed case study constraint-based structure learning is applied to derive hypotheses of causal relationships between four prominent modes of atmospheric low-frequency variability in boreal winter including the Western Pacific Oscillation (WPO), Eastern Pacific Oscillation (EPO), Pacific–North America (PNA) pattern, and North Atlantic Oscillation (NAO). The results are shown in the form of static and temporal independence graphs also known as Bayesian Networks. It is found that WPO and EPO are nearly indistinguishable from the cause–effect perspective as strong simultaneous coupling is identified between the two. In addition, changes in the state of EPO (NAO) may cause changes in the state of NAO (PNA) approximately 18 (3–6) days later. These results are not only consistent with previous findings on dynamical processes connecting different low-frequency modes (e.g., interaction between synoptic and low-frequency eddies) but also provide the basis for formulating new hypotheses regarding the time scale and temporal sequencing of dynamical processes responsible for these connections. Last, the authors propose to use structure learning for climate networks, which are currently based primarily on correlation analysis. While correlation-based climate networks focus on similarity between nodes, independence graphs would provide an alternative viewpoint by focusing on information flow in the network.

Long-term temperature and precipitation trends at the Coweeta Hydrologic Laboratory, Otto, North Carolina, USA

Coweeta Hydrologic Laboratory, located in western North Carolina, USA, is a 2,185 ha basin wherein forest climate monitoring and watershed experimentation began in the early 1930s. An extensive climate and hydrologic network has facilitated research for over 75 years. Our objectives in this paper were to describe the monitoring network, present long-term air temperature and precipitation data, and analyze the temporal variation in the long-term temperature and precipitation record. We found that over the period of record: (1) air temperatures have been increasing significantly since the late 1970s, (2) drought severity and frequency have increased with time, and (3) the precipitation distribution has become more extreme over time. We discuss the implications of these trends within the context of regional and global climate change and forest health.

Collective behaviour of climate indices in the North Pacific air—sea system and its potential relationships with decadal climate changes

A climate network of six climate indices of the North Pacific air—sea system is constructed during the period of 1948–2009. In order to find out the inherent relationship between the intrinsic mechanism of climate index network and the important climate shift, the synchronization behaviour and the coupling behaviour of these indices are investigated. Results indicate that climate network synchronization happened around the beginning of the 1960s, in the middle of the 1970s and at the beginnings of the 1990s and the 2000s separately. These synchronization states were always followed by the decrease of the coupling coefficient. Each synchronization of the network was well associated with the abrupt phase or trend changes of annually accumulated abnormal values of North Pacific sea-surface temperature and 500-hPa height, among which the one that happened in the middle of the 1970s is the most noticeable climate shift. We can also obtain this mysterious shift from the first mode of the empirical orthogonal function of six indices. That is to say, abrupt climate shift in North Pacific air—sea system is not only shown by the phase or trend changes of climate indices, but also might be indicated by the synchronizing and the coupling of climate indices. Furthermore, at the turning point of 1975, there are also abrupt correlation changes in the yearly mode of spatial degree distribution of the sea surface temperature and 500-hPa height in the region of the North Pacific, which further proves the probability of climate index synchronization and coupling shift in air—sea systems.

Estimation of surface albedo and directional reflectance from Moderate Resolution Imaging Spectroradiometer (MODIS) observations

Land surface albedo is one of the key geophysical variables controlling the surface radiation budget. In recent years, land surface albedo products have been generated using data from various satellites. However, some problems exist in those products due to either the failure of the current retrieving procedures resulting from persistent clouds and/or abrupt surface changes, or the reduced temporal or spatial coverage, which may limit their applications. Rapidly generated albedo products that help reduce the impacts of cloud contamination and improve the capture of events such as ephemeral snow and vegetation growth are in demand.In this study, we propose a method for estimating the land surface albedo from Moderate Resolution Imaging Spectroradiometer (MODIS) data using a short temporal window. Instead of executing the atmospheric correction first and then fitting the surface reflectance in the current MODIS albedo procedure, the atmospheric properties (e.g., aerosol optical depth) and surface properties (e.g., surface bidirectional reflectance) were estimated simultaneously. Validations were carried out using various data sources including ground measurements (e.g., from the Surface Radiation (SURFRAD) Network and Greenland Climate Network (GC-Net)) and MODIS AERONET-based Surface Reflectance Validation Network (MODASRVN) data. The results showed comparable albedo estimates with both MODIS data and ground measurements, and the MODASRVN instantaneous surface reflectance was in good agreement with the reflectance estimation from our method. Aerosol optical depth (AOD) retrievals over SURFRAD and MODASRVN sites were also compared with ground measurements. Validation results showed estimation accuracies similar to those of MODIS aerosol products.

Very high resolution regional climate model simulations over Greenland: Identifying added value

This study presents two simulations of the climate over Greenland with the regional climate model (RCM) HIRHAM5 at 0.05° and 0.25° resolution driven at the lateral boundaries by the ERA‐Interim reanalysis for the period 1989–2009. These simulations are validated against observations from meteorological stations (Danish Meteorological Institute) at the coast and automatic weather stations on the ice sheet (Greenland Climate Network). Generally, the temperature and precipitation biases are small, indicating a realistic simulation of the climate over Greenland that is suitable to drive ice sheet models. However, the bias between the simulations and the few available observations does not reduce with higher resolution. This is partly explained by the lack of observations in regions where the higher resolution is expected to improve the simulated climate. The RCM simulations show that the temperature has increased the most in the northern part of Greenland and at lower elevations over the period 1989–2009. Higher resolution increases the relief variability in the model topography and causes the simulated precipitation to be larger on the coast and smaller over the main ice sheet compared to the lower‐resolution simulation. The higher‐resolution simulation likely represents the Greenlandic climate better, but the lack of observations makes it difficult to validate fully. The detailed temperature and precipitation fields that are generated with the higher resolution are recommended for producing adequate forcing fields for ice sheet models, particularly for their improved simulation of the processes occurring at the steep margins of the ice sheet.

Modeling clear‐sky solar radiation across a range of elevations in Hawai‘i: Comparing the use of input parameters at different temporal resolutions

Estimates of clear sky global solar irradiance using the parametric model SPCTRAL2 were tested against clear sky radiation observations at four sites in Hawai‘i using daily, mean monthly, and 1 year mean model parameter settings. Atmospheric parameters in SPCTRAL2 and similar models are usually set at site‐specific values and are not varied to represent the effects of fluctuating humidity, aerosol amount and type, or ozone concentration, because time‐dependent atmospheric parameter estimates are not available at most sites of interest. In this study, we sought to determine the added value of using time dependent as opposed to fixed model input parameter settings. At the AERONET site, Mauna Loa Observatory (MLO) on the island of Hawai‘i, where daily measurements of atmospheric optical properties and hourly solar radiation observations are available, use of daily rather than 1 year mean aerosol parameter values reduced mean bias error (MBE) from 18 to 10 W m−2 and root mean square error from 25 to 17 W m−2. At three stations in the HaleNet climate network, located at elevations of 960, 1640, and 2590 m on the island of Maui, where aerosol‐related parameter settings were interpolated from observed values for AERONET sites at MLO (3397 m) and Lāna‘i (20 m), and precipitable water was estimated using radiosonde‐derived humidity profiles from nearby Hilo, the model performed best when using constant 1 year mean parameter values. At HaleNet Station 152, for example, MBE was 18, 10, and 8 W m−2 for daily, monthly, and 1 year mean parameters, respectively.

Observed and Modeled Twentieth-Century Spatial and Temporal Patterns of Selected Agro-Climate Indices in North America

Abstract The authors examine recent changes in three agro-climate indices (frost days, thermal time, and heat stress index) in North America (centered around the continental United States) using observations from a historical climate network and an ensemble of 17 global climate models (GCMs) from the Fourth Assessment Report of the Intergovernmental Panel on Climate Change (IPCC AR4). Agro-climate indices provide the basis for analyzing agricultural time series that are unbiased by long-term technological intervention. Observations from the last 60 years (1951–2010) confirm conclusions of previous studies showing continuing declines in the number of frost days and increases in thermal time. Increases in heat stress are largely confined to the western half of the continent. The authors do not observe accelerating agro-climate warming trends in the most recent decade of observations. The spatial variability of the temporal trends in GCMs is lower compared to the observed patterns, which still show some regional cooling trends. GCM skill, defined as the ability to reproduce observed patterns (i.e., correlation and error) and variability, is highest for frost days and lowest for heat stress patterns. Individual GCM skill is incorporated into two model weighting schemes to gauge their ability to reduce predictive uncertainty for agro-climate indices. The two weighted GCM ensembles do not substantially improve results compared to the unweighted ensemble mean. The lack of agreement between simulated and observed heat stress is relatively robust with respect to how the heuristic is defined and appears to reflect a weakness in the ability of this last generation of GCMs to reproduce this impact-relevant aspect of the climate system. However, it remains a question for future work as to whether the discrepancies between observed and simulated trends primarily reflect fundamental errors in model physics or an incomplete treatment of relevant regional climate forcings.

Calibration and evaluation of a high-resolution surface mass-balance model for Paakitsoq, West Greenland

AbstractModelling the hydrology of the Greenland ice sheet, including the filling and drainage of supraglacial lakes, requires melt inputs generated at high spatial and temporal resolution. Here we apply a high spatial (100 m) and temporal (1 hour) mass-balance model to a 450 km2 subset of the Paakitsoq region, West Greenland. The model is calibrated by adjusting the values for parameters of fresh snow density, threshold temperature for solid/liquid precipitation and elevation-dependent precipitation gradient to minimize the error between modelled output and surface height and albedo measurements from three Greenland Climate Network stations for the mass-balance years 2000/01 and 2004/05. Bestfit parameter values are consistent between the two years at 400 kg m-3, 2°C and +14% (100 m)-1, respectively. Model performance is evaluated, first, by comparing modelled snow and ice distribution with that derived from Landsat-7 ETM+ satellite imagery using normalized-difference snow index classification and supervised image thresholding; and second, by comparing modelled albedo with that retrieved from the MODIS sensor M0D10A1 product. Calculation of mass-balance components indicates that 6% of surface meltwater and rainwater refreezes in the snowpack and does not become runoff, such that refreezing accounts for 31% of the net accumulation.

The Impact of Climate Modes on Summer Temperature and Precipitation of Darwin, Australia, 1870-2011

Monthly mean summer (DJF) temperature and precipitation from Global Historical Climate Network (GHCN-V3) for the period of 1870-2011, are analyzed to assess the role of teleconnections on climate of Darwin, Australia. Indices of El Nino-Southern Oscillation (ENSO), Antarctic Oscillation (AAO), Pacific Decadal Oscillation (PDO), North Atlantic Oscillation (NAO), Arctic Oscillation (AO), and Pacific North American Oscillation (PNA) are extracted from monthly means and compared with climatic data of Darwin. Most of these climate modes are shown to have a strong influence on the monthly mean summer temperature and precipitation. ENSO is shown to have a positive relationship with the amount of precipitation received and a negative relationship with the temperature. Where an El Nino event produces warmer drier conditions and a La Nina event produces colder wetter conditions. The AAO is shown to cause cold and dry conditions during the positive phase and warm and wet conditions during the negative phase. The PDO is shown to cause El Nino like condition during the positive phase causing warmer, drier weather, and La Nina like conditions during the negative phase causing cooler, wetter weather. Through the analysis it is also shown that the NAO, AO, and PNA have little effect on the temperature and precipitation patterns of Darwin.

Development of a Unified Land Model for Prediction of Surface Hydrology and Land–Atmosphere Interactions

AbstractA unified land model (ULM) is described that combines the surface flux parameterizations in the Noah land surface model (used in most of NOAA’s coupled weather and climate models) with the Sacramento Soil Moisture Accounting model (Sac; used for hydrologic prediction within the National Weather Service). The motivation was to develop a model that has a history of strong hydrologic performance while having the ability to be run in the coupled land–atmosphere environment. ULM takes the vegetation, snow model, frozen soil, and evapotranspiration schemes from Noah and merges them with the soil moisture accounting scheme from Sac. ULM surface fluxes, soil moisture, and streamflow simulations were evaluated through comparisons with observations from the Ameriflux (surface flux), Illinois Climate Network (soil moisture), and Model Parameter Estimation Experiment (MOPEX; streamflow) datasets. Initially, a priori parameters from Sac and Noah were used, which resulted in ULM surface flux simulations that were comparable to those produced by Noah (Sac does not predict surface energy fluxes). ULM with the a priori parameters had streamflow simulation skill that was generally similar to Sac’s, although it was slightly better (worse) for wetter (more arid) basins. ULM model performance using a set of parameters identified via a Monte Carlo search procedure lead to substantial improvements relative to the a priori parameters. A scheme for transfer of parameters from streamflow simulations to nearby flux and soil moisture measurement points was also evaluated; this approach did not yield conclusive improvements relative to the a priori parameters.

Discerning connectivity from dynamics in climate networks

Abstract. The bias due to dynamical memory (serial correlations) in an association/dependence measure (absolute cross-correlation) is demonstrated in model data and identified in time series of meteorological variables used for construction of climate networks. Accounting for such bias in inferring links of the climate network markedly changes the network topology and allows to observe previously hidden phenomena in climate network evolution.

Community-based anomaly detection in evolutionary networks

Networks of dynamic systems, including social networks, the World Wide Web, climate networks, and biological networks, can be highly clustered. Detecting clusters, or communities, in such dynamic networks is an emerging area of research; however, less work has been done in terms of detecting community-based anomalies. While there has been some previous work on detecting anomalies in graph-based data, none of these anomaly detection approaches have considered an important property of evolutionary networks--their community structure. In this work, we present an approach to uncover community-based anomalies in evolutionary networks characterized by overlapping communities. We develop a parameter-free and scalable algorithm using a proposed representative-based technique to detect all six possible types of community-based anomalies: grown, shrunken, merged, split, born, and vanished communities. We detail the underlying theory required to guarantee the correctness of the algorithm. We measure the performance of the community-based anomaly detection algorithm by comparison to a non---representative-based algorithm on synthetic networks, and our experiments on synthetic datasets show that our algorithm achieves a runtime speedup of 11---46 over the baseline algorithm. We have also applied our algorithm to two real-world evolutionary networks, Food Web and Enron Email. Significant and informative community-based anomaly dynamics have been detected in both cases.

Emergence of El Niño as an Autonomous Component in the Climate Network

We construct and analyze a climate network which represents the interdependent structure of the climate in different geographical zones and find that the network responds in a unique way to El Niño events. Analyzing the dynamics of the climate network shows that when El Niño events begin, the El Niño basin partially loses its influence on its surroundings. After typically three months, this influence is restored while the basin loses almost all dependence on its surroundings and becomes autonomous. The formation of an autonomous basin is the missing link to understand the seemingly contradicting phenomena of the afore-noticed weakening of the interdependencies in the climate network during El Niño and the known impact of the anomalies inside the El Niño basin on the global climate system.

CHANGE: Climate and Hydrology Academic Network for Governance and the Environment

No abstract available.

CHANGE: Climate and Hydrology Academic Network for Governance and the Environment

CHANGE: Climate and Hydrology Academic Network for Governance and the Environment

A New Homogenized Climate Division Precipitation Dataset for Analysis of Climate Variability and Climate Change

AbstractA new homogeneous climate division monthly precipitation dataset [based on full network estimated precipitation (FNEP)] was created as an alternative to the National Climatic Data Center (NCDC) climate division dataset. These alternative climate division monthly precipitation values were estimated using an equal-weighted average of Cooperative Observer Program stations that contained serially complete time series. Missing station observations were estimated by a procedure that was optimized through testing on U.S. Historical Climate Network stations. Inhomogeneities in the NCDC dataset arise from two principal causes. The pre-1931 estimation of NCDC climate division monthly precipitation from statewide averages led to a significant time series discontinuity in several climate divisions. From 1931 to the present, NCDC climate division averages have been calculated from a subset of available station data within each climate division, and temporal changes in the location of available stations have caused artificial changes in the time series. The FNEP climate division dataset is recommended over the NCDC dataset for studies involving climate trends or long-term climate variability. According to the FNEP data, the 1895–2009 linear precipitation trend is positive across most of the United States, and trends exceed 10% per century across the southern plains and the Corn Belt. Remaining inhomogeneities from changes in gauge technology and station location may be responsible for an artificial trend of 1%–3% per century.

Investigating the topology of interacting networks

Network theory provides various tools for investigating the structural or functional topology of many complex systems found in nature, technology and society. Nevertheless, it has recently been realised that a considerable number of systems of interest should be treated, more appropriately, as interacting networks or networks of networks. Here we introduce a novel graph-theoretical framework for studying the interaction structure between subnetworks embedded within a complex network of networks. This framework allows us to quantify the structural role of single vertices or whole subnetworks with respect to the interaction of a pair of subnetworks on local, mesoscopic and global topological scales. Climate networks have recently been shown to be a powerful tool for the analysis of climatological data. Applying the general framework for studying interacting networks, we introduce coupled climate subnetworks to represent and investigate the topology of statistical relationships between the fields of distinct climatological variables. Using coupled climate subnetworks to investigate the terrestrial atmosphere's three-dimensional geopotential height field uncovers known as well as interesting novel features of the atmosphere's vertical stratification and general circulation. Specifically, the new measure "cross-betweenness" identifies regions which are particularly important for mediating vertical wind field interactions. The promising results obtained by following the coupled climate subnetwork approach present a first step towards an improved understanding of the Earth system and its complex interacting components from a network perspective.

Inferring long memory processes in the climate network via ordinal pattern analysis

We use ordinal patterns and symbolic analysis to construct global climate networks and uncover long- and short-term memory processes. Data analyzed are the monthly averaged surface air temperature (SAT field), and the results suggest that the time variability of the SAT field is determined by patterns of oscillatory behavior that repeat from time to time, with a periodicity related to intraseasonal oscillations and to El Niño on seasonal-to-interannual time scales.

The identification of distinct patterns in California temperature trends

Regional changes in California surface temperatures over the last 80 years are analyzed using station data from the US Historical Climate Network and the National Weather Service Cooperative Network. Statistical analyses using annual and seasonal temperature data over the last 80 years show distinctly different spatial and temporal patterns in trends of maximum temperature (Tmax) compared to trends of minimum temperature (Tmin). For trends computed between 1918 and 2006, the rate of warming in Tmin is greater than that of Tmax. Trends computed since 1970 show an amplified warming rate compared to trends computed from 1918, and the rate of warming is comparable between Tmin and Tmax. This is especially true in the southern deserts, where warming trends during spring (March–May) are exceptionally large. While observations show coherent statewide positive trends in Tmin, trends in Tmax vary on finer spatial and temporal scales. Accompanying the observed statewide warming from 1970 to 2006, regional cooling trends in Tmax are observed during winter and summer. These signatures of regional temperature change suggest that a collection of different forcing mechanisms or feedback processes must be present to produce these responses.

Distribution of Landscape Types in the Global Historical Climatology Network

Abstract The Global Historical Climate Network version 2 (GHCNv.2) surface temperature dataset is widely used for reconstructions such as the global average surface temperature (GAST) anomaly. Because land use and land cover (LULC) affect temperatures, it is important to examine the spatial distribution and the LULC representation of GHCNv.2 stations. Here, nightlight imagery, two LULC datasets, and a population and cropland historical reconstruction are used to estimate the present and historical worldwide occurrence of LULC types and the number of GHCNv.2 stations within each. Results show that the GHCNv.2 station locations are biased toward urban and cropland (>50% stations versus 18.4% of the world’s land) and past century reclaimed cropland areas (35% stations versus 3.4% land). However, widely occurring LULC such as open shrubland, bare, snow/ice, and evergreen broadleaf forests are underrepresented (14% stations versus 48.1% land), as well as nonurban areas that have remained uncultivated in the past century (14.2% stations versus 43.2% land). Results from the temperature trends over the different landscapes confirm that the temperature trends are different for different LULC and that the GHCNv.2 stations network might be missing on long-term larger positive trends. This opens the possibility that the temperature increases of Earth’s land surface in the last century would be higher than what the GHCNv.2-based GAST analyses report.