Error characteristics of high resolution regional climate models over the Alpine area

This study describes typical error ranges of high resolution regional climate models operated over complex orography and investigates the scale-dependence of these error ranges. The results are valid primarily for the European Alpine region, but to some extent they can also be transferred to other orographically complex regions of the world. We investigate the model errors by evaluating a set of 62 one-year hindcast experiments for the year 1999 with four different regional climate models. The analysis is conducted for the parameters mean sea level pressure, air temperature (mean, minimum and maximum) and precipitation (mean, frequency and intensity), both as an area average over the whole modeled domain (the “Greater Alpine Region”, GAR) and in six subregions. The subregional seasonal error ranges, defined as the interval between the 2.5th percentile and the 97.5th percentile, lie between ?3.2 and +2.0 K for temperature and between ?2.0 and +3.1 mm/day (?45.7 and +94.7%) for precipitation, respectively. While the temperature error ranges are hardly broadened at smaller scales, the precipitation error ranges increase by 28%. These results demonstrate that high resolution RCMs are applicable in relatively small scale climate impact studies with a comparable quality as on well investigated larger scales as far as temperature is concerned. For precipitation, which is a much more demanding parameter, the quality is moderately degraded on smaller scales.     

Comparison of artificial intelligence techniques for prediction of soil temperatures in Turkey

Soil temperature is a meteorological data directly affecting the formation and development of plants of all kinds. Soil temperatures are usually estimated with various models including the artificial neural networks (ANNs), adaptive neuro-fuzzy inference system (ANFIS), and multiple linear regression (MLR) models. Soil temperatures along with other climate data are recorded by the Turkish State Meteorological Service (MGM) at specific locations all over Turkey. Soil temperatures are commonly measured at 5-, 10-, 20-, 50-, and 100-cm depths below the soil surface. In this study, the soil temperature data in monthly units measured at 261 stations in Turkey having records of at least 20 years were used to develop relevant models. Different input combinations were tested in the ANN and ANFIS models to estimate soil temperatures, and the best combination of significant explanatory variables turns out to be monthly minimum and maximum air temperatures, calendar month number, depth of soil, and monthly precipitation. Next, three standard error terms (mean absolute error (MAE, °C), root mean squared error (RMSE, °C), and determination coefficient (R ²)) were employed to check the reliability of the test data results obtained through the ANN, ANFIS, and MLR models. ANFIS (RMSE 1.99; MAE 1.09; R ² 0.98) is found to outperform both ANN and MLR (RMSE 5.80, 8.89; MAE 1.89, 2.36; R ² 0.93, 0.91) in estimating soil temperature in Turkey.

     

Hourly air temperature driven using multi-layer perceptron and radial basis function networks in arid and semi-arid regions

This study employed two artificial neural network (ANN) models, including multi-layer perceptron (MLP) and radial basis function (RBF), as data-driven methods of hourly air temperature at three meteorological stations in Fars province, Iran. MLP was optimized using the Levenberg–Marquardt (MLP_LM) training algorithm with a tangent sigmoid transfer function. Both time series (TS) and randomized (RZ) data were used for training and testing of ANNs. Daily maximum and minimum air temperatures (MM) and antecedent daily maximum and minimum air temperatures (AMM) constituted the input for ANNs. The ANN models were evaluated using the root mean square error (RMSE), the coefficient of determination (R ²) and the mean absolute error. The use of AMM led to a more accurate estimation of hourly temperature compared with the use of MM. The MLP-ANN seemed to have a higher estimation efficiency than the RBF ANN. Furthermore, the ANN testing using randomized data showed more accurate estimation. The RMSE values for MLP with RZ data using daily maximum and minimum air temperatures for testing phase were equal to 1.2°C, 1.8°C, and 1.7°C, respectively, at Arsanjan, Bajgah, and Kooshkak stations. The results of this study showed that hourly air temperature driven using ANNs (proposed models) had less error than the empirical equation.     

Estimation of evapotranspiration based on only air temperature data using artificial neural networks for a subtropical climate in Iran

This paper examines the potential for the use of artificial neural networks (ANNs) to estimate the reference crop evapotranspiration (ET₀) based on air temperature data under humid subtropical conditions on the southern coast of the Caspian Sea situated in the north of Iran. The input variables for the networks were the maximum and minimum air temperature and extraterrestrial radiation. The temperature data were obtained from eight meteorological stations with a range of latitude, longitude, and elevation throughout the study area. A comparison of the estimates provided by the ANNs and by Hargreaves equation was also conducted. The FAO-56 Penman–Monteith model was used as a reference model for assessing the performance of the two approaches. The results of this study showed that ANNs using air temperature data successfully estimated the daily ET₀ and that the ANNs with an R ² of 0.95 and a root mean square error (RMSE) of 0.41 mm day^?1 simulated ET₀ better than the Hargreaves equation, which had an R ² of 0.91 and a RMSE of 0.51 mm day^?1.     

Snowpack and runoff response to climate change in Owens Valley and Mono Lake watersheds

Precipitation from the Eastern Sierra Nevada watersheds of Owens Lake and Mono Lake is one of the main water sources for Los Angeles’ over 4 million people, and plays a major role in the ecology of Mono Lake and of these watersheds. We use the Variable Infiltration Capacity (VIC) hydrologic model at daily time scale, forced by climate projections from 16 global climate models under greenhouse gas emissions scenarios B1 and A2, to evaluate likely hydrologic responses in these watersheds for 1950–2099. Comparing climate in the latter half of the 20th Century to projections for 2070–2099, we find that all projections indicate continued temperature increases, by 2–5 °C, but differ on precipitation changes, ranging from ?24 % to +56 %. As a result, the fraction of precipitation falling as rain is projected to increase, from a historical 0.19 to a range of 0.26–0.52 (depending on the GCM and emission scenario), leading to earlier timing of the annual hydrograph’s center, by a range of 9–37 days. Snowpack accumulation depends on temperature and even more strongly on precipitation due to the high elevation of these watersheds (reaching 4,000 m), and projected changes for April 1 snow water equivalent range from ?67 % to +9 %. We characterize the watershed’s hydrologic response using variables integrated in space over the entire simulated area and aggregated in time over 30-year periods. We show that from the complex dynamics acting at fine time scales (seasonal and sub-seasonal) simple dynamics emerge at this multi-year time scale. Of particular interest are the dynamic effects of temperature. Warming anticipates hydrograph timing, by raising the fraction of precipitation falling as rain, reducing the volume of snowmelt, and initiating snowmelt earlier. This timing shift results in the depletion of soil moisture in summer, when potential evapotranspiration is highest. Summer evapotranspiration losses are limited by soil moisture availability, and as a result the watershed’s water balance at the annual and longer scales is insensitive to warming. Mean annual runoff changes at base-of-mountain stations are thus strongly determined by precipitation changes.     

A more accurate scheme for calculating Earth’s skin temperature

The theoretical framework of the vertical discretization of a ground column for calculating Earth’s skin temperature is presented. The suggested discretization is derived from the evenly heat-content discretization with the optimal effective thickness for layer-temperature simulation. For the same level number, the suggested discretization is more accurate in skin temperature as well as surface ground heat flux simulations than those used in some state-of-the-art models. A proposed scheme (“op(3,2,0)”) can reduce the normalized root–mean–square error (or RMSE/STD ratio) of the calculated surface ground heat flux of a cropland site significantly to 2% (or 0.9 W m^?2), from 11% (or 5 W m^?2) by a 5-layer scheme used in ECMWF, from 19% (or 8 W m^?2) by a 5-layer scheme used in ECHAM, and from 74% (or 32 W m^?2) by a single-layer scheme used in the UCLA GCM. Better accuracy can be achieved by including more layers to the vertical discretization. Similar improvements are expected for other locations with different land types since the numerical error is inherited into the models for all the land types. The proposed scheme can be easily implemented into state-of-the-art climate models for the temperature simulation of snow, ice and soil.     

Feedback process responsible for the suppression of ENSO activity during the mid-Holocene

Using the output of 12 models from the Paleoclimate Modeling Intercomparison Project Phase 3, we investigate the feedback process responsible for changes in El Niño-Southern Oscillation activity during the mid-Holocene based on a linear stability index (Bjerknes stability index; BJ index) analysis. The multi-model ensemble mean (MME) variance of the Niño-3.4 index (sea surface temperature anomalies averaged over 5°S–5°N, 170°–120°W) simulated for 6000 years ago (6 kya) was 13% lower than that for the pre-industrial era (0 kya), while changes in the MME BJ index were negligible. This is due to a balance between enhanced damping by anomalous thermal advection by mean currents (MA) and enhanced positive thermocline feedback (TH). Seven of the models show that MME variance of the Niño-3.4 and BJ indexes for the 6 kya run is 21 and 70% lower, respectively, than the 0 kya run. However, two models show the opposite change. Interestingly, MA in both model groups increases, especially due to the mean meridional current associated with enhanced trade winds, indicating a robust mechanism. The opposite tendency between the two groups is mainly due to large TH in the second group 6 kya, as a result of enhanced air-sea coupling and strongly reduced ocean stratification due to subsurface warming, which led to increased sensitivity of the zonal thermocline contrast to surface zonal wind stress.     

The contribution of snow condition trends to future ground climate

Global climate models predict that terrestrial northern high-latitude snow conditions will change substantially over the twenty-first century. Results from a Community Climate System Model simulation of twentieth and twenty-first (SRES A1B scenario) century climate show increased winter snowfall (+10–40%), altered maximum snow depth (?5 ± 6 cm), and a shortened snow-season (?14 ± 7 days in spring, +20 ± 9 days in autumn). By conducting a series of prescribed snow experiments with the Community Land Model, we isolate how trends in snowfall, snow depth, and snow-season length affect soil temperature trends. Increasing snowfall, by countering the snowpack-shallowing influence of warmer winters and shorter snow seasons, is effectively a soil warming agent, accounting for 10–30% of total soil warming at 1 m depth and ~16% of the simulated twenty-first century decline in near-surface permafrost extent. A shortening snow season enhances soil warming due to increased solar absorption whereas a shallowing snowpack mitigates soil warming due to weaker winter insulation from cold atmospheric air. Snowpack deepening has comparatively less impact due to saturation of snow insulative capacity at deeper snow depths. Snow depth and snow-season length trends tend to be positively related, but their effects on soil temperature are opposing. Consequently, on the century timescale the net change in snow state can either amplify or mitigate soil warming. Snow state changes explain less than 25% of total soil temperature change by 2100. However, for the latter half of twentieth century, snow state variations account for as much as 50–100% of total soil temperature variations.     

A model of the ground surface temperature for micrometeorological analysis

Micrometeorological models at various scales require ground surface temperature, which may not always be measured in sufficient spatial or temporal detail. There is thus a need for a model that can calculate the surface temperature using only widely available weather data, thermal properties of the ground, and surface properties. The vegetated/permeable surface energy balance (VP-SEB) model introduced here requires no a priori knowledge of soil temperature or moisture at any depth. It combines a two-layer characterization of the soil column following the heat conservation law with a sinusoidal function to estimate deep soil temperature, and a simplified procedure for calculating moisture content. A physically based solution is used for each of the energy balance components allowing VP-SEB to be highly portable. VP-SEB was tested using field data measuring bare loess desert soil in dry weather and following rain events. Modeled hourly surface temperature correlated well with the measured data (r ² = 0.95 for a whole year), with a root-mean-square error of 2.77 K. The model was used to generate input for a pedestrian thermal comfort study using the Index of Thermal Stress (ITS). The simulation shows that the thermal stress on a pedestrian standing in the sun on a fully paved surface, which may be over 500 W on a warm summer day, may be as much as 100 W lower on a grass surface exposed to the same meteorological conditions.     

Temperature and thermal growing season variations along elevational gradients on a sub-alpine,temperate China

Fully and accurately studying temperature variations in montane areas are conducive to a better understanding of climate modeling and climate-growth relationships on regional scales. To explore the spatio-temporal changes in air and soil temperatures and their relationship in montane areas, on-site monitoring over 2 years (2015 and 2016) was conducted at nine different elevations from 2040 to 2740 m a.s.l. on Luya Mountain in the semiarid region of China. The results showed that the annual mean of air temperature lapse rate (ATLR) was 0.67 °C/100 m. ATLR varied obviously in different months within a range of 0.57~0.79 °C/100 m. The annual mean of the soil temperature lapse rate (STLR) was 0.48 °C/100 m. Seasonally, monthly mean soil temperature did not show a consistent pattern with regard to elevation. The relationships between air and soil temperatures showed piecewise changes. Soil was decoupled from the air temperature in cold winter and early spring. The parameters of the growing season based on the two temperature types had no corresponding relations, and seasonal mean of soil temperature showed the smallest value at mid-elevation rather than in the treeline ecotone. Based on these changes, our results emphasized that altitudinal and seasonal variability caused by local factors (such as snow cover and soil moisture) should be taken into full consideration in microclimate extrapolation and treeline prediction in montane areas, especially in relation to soil temperature.     

Estimation of ground temperatures in permafrost regions of the Qinghai-Tibetan Plateau from climatic variables

Ground temperature is an important factor influencing ground source heat pumps, ground energy storage systems, land-atmosphere processes, and ecosystem dynamics. This paper presents an accurate development model (DM) based on a segment function: it can derive ground temperatures in permafrost regions of the Qinghai-Tibetan Plateau (QTP) from air temperature in case of shallow soil depths and without using air temperature data in case of deep soil depths. Here, we applied this model to simulate the active layer and permafrost ground temperature at the Tanggula observation station. The DM results were compared with those from the artificial neural network (ANN), support vector machine (SVM), and multiple linear regressions (MLR) models, which were based on climatic variables from prior models and on ground temperatures derived from observations at different depths. The results revealed that the effect of air temperature on simulated ground temperatures decreased with increasing depth; moreover, ground temperatures fluctuated greatly within the shallow layers but remained rather stable with deeper layers. The results also indicated that the DM has the best performance for the estimation of soil temperature compared to the MLR, SVM, and ANN models. Finally, we obtained the three average statistics indexes, i.e., mean absolute error (MAE), root mean square error (RMSE), and the normalized standard error (NSEE) at TGL site: they were equal to 0.51 °C, 0.63 °C, and 0.15 °C for the ground temperature of active layer and to 0.08 °C, 0.09 °C, and 0.07 °C for the permafrost temperature.     

Analysis and estimation of the effects of missing values on the calculation of monthly temperature indices

Long and complete climatic data series are a fundamental resource for scientific research on climate change. Data quality is important, and missing value or data gap management is a key process that must be dealt with carefully to produce reliable datasets. Although a large variety of techniques are available for gap-filling, a widespread strategy is to consider a dataset reliable if the rate of missing data is below a given threshold. However this strategy varies from study to study. The aim of this paper is to analyze the impact of missing daily values on the estimation of monthly average temperature indices. The relationship between the error of the estimate and the presence of random or consecutive missing values, as well as data series autocorrelation is also analyzed. A theoretical, a linear and a nonlinear model to estimate the maximum error at the 95 % confidence interval are tested on data series provided by national and worldwide networks of stations. Consecutive missing values have an important effect on error estimation due to autocorrelation of temperature data series. On our dataset, the mean and standard deviation of the error for five consecutive missing values (0.27?±?0.05 °C) on a normalized daily series (σ?=?1) was higher than for five random missing values (0.14?±?0.006 °C). A nonlinear model taking into account the number of consecutive missing values is able to estimate the error and its performance is less affected by the presence of consecutive missing values than the other proposed models.     

The 20th century cooling trend over the southeastern United States

Portions of the southern and southeastern United States, primarily Mississippi, Alabama, and Georgia, have experienced century-long (1895–2007) downward air temperature trends that occur in all seasons. Superimposed on them are shifts in mean temperatures on decadal scales characterized by alternating warm (1930s–1940s, 1990s) and cold (1900s; 1960s–1970s) regimes. Regional atmospheric circulation and SST teleconnection indices, station-based cloud cover and soil moisture (Palmer drought severity index) data are used in stepwise multiple linear regression models. These models identify predictors linked to observed winter, summer, and annual Southeastern air temperature variability, the observed variance (r²) they explain, and the resulting prediction and residual time series. Long-term variations and trends in tropical Pacific sea temperatures, cloud cover, soil moisture and the North Atlantic and Arctic oscillations account for much of the air temperature downtrends. Soil moisture and cloud cover are the primary predictors of 59.6 % of the observed summer temperature variance. While the teleconnections, cloud cover and moisture data account for some of the annual and summer Southeastern cooling trend, large significant downward trending residuals remain in winter and summer. Comparison is made to the northeastern United States where large twentieth century upward air temperature trends are driven by cloud cover increases and Atlantic Multidecadal Oscillation (AMO) variability. Differences between the Northeastern warming and the Southeastern cooling trends in summer are attributable in part to the differing roles of cloud cover, soil moisture, the Arctic Oscillation and the AMO on air temperatures of the 2 regions.     

Monthly rainfall prediction using wavelet regression and neural network: an analysis of 1901–2002 data,Assam, India

Rainfall is a principal element of the hydrological cycle and its variability is important from both the scientific as well as practical point of view. Wavelet regression (WR) technique is proposed and developed to analyze and predict the rainfall forecast in this study. The WR model is improved combining two methods, discrete wavelet transform and linear regression model. This study uses rainfall data from 21 stations in Assam, India over 102 years from 1901 to 2002. The calibration and validation performance of the models is evaluated with appropriate statistical methods. The root mean square errors (RMSE), N-S index, and correlation coefficient (R) statistics were used for evaluating the accuracy of the WR models. The accuracy of the WR models was then compared with those of the artificial neural networks (ANN) models. The results of monthly rainfall series modeling indicate that the performances of wavelet regression models are found to be more accurate than the ANN models.     

Statistical estimation of high-resolution surface air temperature from MODIS over the Yangtze River Delta,China

High-resolution surface air temperature data are critical to regional climate modeling in terms of energy balance, urban climate change, and so on. This study demonstrates the feasibility of using Moderate Resolution Imaging Spectroradiometer (MODIS) land surface temperature (LST) to estimate air temperature at a high resolution over the Yangtze River Delta region, China. It is found that daytime LST is highly correlated with maximum air temperature, and the linear regression coefficients vary with the type of land surface. The air temperature at a resolution of 1 km is estimated from the MODIS LST with linear regression models. The estimated air temperature shows a clear spatial structure of urban heat islands. Spatial patterns of LST and air temperature differences are detected, indicating maximum differences over urban and forest regions during summer. Validations are performed with independent data samples, demonstrating that the mean absolute error of the estimated air temperature is approximately 2.5°C, and the uncertainty is about 3.1°C, if using all valid LST data. The error is reduced by 0.4°C (15%) if using best-quality LST with errors of less than 1 K. The estimated high-resolution air temperature data have great potential to be used in validating high-resolution climate models and other regional applications.     

DEM-based delineation for improving geostatistical interpolation of rainfall in mountainous region of Central Himalayas,India

In mountainous region with heterogeneous topography, the geostatistical modeling of the rainfall using global data set may not confirm to the intrinsic hypothesis of stationarity. This study was focused on improving the precision of the interpolated rainfall maps by spatial stratification in complex terrain. Predictions of the normal annual rainfall data were carried out by ordinary kriging, universal kriging, and co-kriging, using 80-point observations in the Indian Himalayas extending over an area of 53,484 km². A two-step spatial clustering approach is proposed. In the first step, the study area was delineated into two regions namely lowland and upland based on the elevation derived from the digital elevation model. The delineation was based on the natural break classification method. In the next step, the rainfall data was clustered into two groups based on its spatial location in lowland or upland. The terrain ruggedness index (TRI) was incorporated as a co-variable in co-kriging interpolation algorithm. The precision of the kriged and co-kriged maps was assessed by two accuracy measures, root mean square error and Chatfield’s percent better. It was observed that the stratification of rainfall data resulted in 5–20 % of increase in the performance efficiency of interpolation methods. Co-kriging outperformed the kriging models at annual and seasonal scale. The result illustrates that the stratification of the study area improves the stationarity characteristic of the point data, thus enhancing the precision of the interpolated rainfall maps derived using geostatistical methods.     

Predictability of NDVI in semi-arid African regions

In semi-arid Africa, rainfall variability is an important issue for ecosystems and agricultural activities. However, due to its discrete nature in time and space, rainfall is difficult to measure, quantify, and predict. In the dry tropics, a good proxy for rainfall is vegetation activity since this parameter is well correlated with rainfall variations. In this study, over 20 years of Normalized Difference Vegetative Index (NDVI) data from the Advanced Very High Resolution Radiometers are used. The goal is to assess the skill of linear statistical models in estimating regional NDVI interannual variability based on ocean and atmospheric fields (but not rainfall) and then to hindcast it with a 1- to 2-month lead-time. Three semi-arid areas of ～150 000 km² located in Western, Southern, and Eastern Tropical Africa are considered for this purpose. The predictors are: the Niño3.4 sea surface temperature index, the main modes of National Center for Environmental Prediction (NCEP) surface temperature variability in a window centered over Africa, and regional-scale indices based on NCEP surface temperatures and atmospheric variables (relative humidity, geopotential heights, and winds). The regional indices, which are physically and statistically robust, are generally asynchronous with the NDVI predictand. The statistical models, based on linear multiple regressions, give significant results, and the correlation between observed and cross-validated NDVI is 0.67 in Southern Africa, 0.76 for the long rains and 0.83 for the short rains in Eastern Africa, and 0.88 in Western Africa. The results have implications for (1) better understanding the role of El Niño/Southern Oscillation in semi-arid Africa, and (2) highlight the importance of regional climate processes for vegetation growth at these scales, notably the role played by the Mediterranean Sea and its influence on the West African monsoon. The predictability of NDVI over these African regions is discussed.     

Air Temperature Estimation with MODIS Data over the Northern Tibetan Plateau

Time series of MODIS land surface temperature(T_s) and normalized difference vegetation index(NDVI) products,combined with digital elevation model(DEM) and meteorological data from 2001 to 2012,were used to map the spatial distribution of monthly mean air temperature over the Northern Tibetan Plateau(NTP). A time series analysis and a regression analysis of monthly mean land surface temperature(T_s) and air temperature(T_a) were conducted using ordinary linear regression(OLR) and geographical weighted regression(GWR). The analyses showed that GWR,which considers MODIS T_s,NDVI and elevation as independent variables,yielded much better results [R_(Adj)~2 0.79; root-mean-square error(RMSE) =0.51℃–1.12℃] associated with estimating T_a compared to those from OLR(R_(Adj)~2= 0.40-0.78; RMSE = 1.60℃–4.38℃).In addition,some characteristics of the spatial distribution of monthly T_a and the difference between the surface and air temperature(T_d) are as follows. According to the analysis of the 0℃ and 10℃ isothermals,T_a values over the NTP at elevations of 4000–5000 m were greater than 10℃ in the summer(from May to October),and T_a values at an elevation of3200 m dropped below 0℃ in the winter(from November to April). T_a exhibited an increasing trend from northwest to southeast. Except in the southeastern area of the NTP,T d values in other areas were all larger than 0℃ in the winter.     

High resolution regional climate model simulations for Germany: Part II—projected climate changes

The projected climate change signals of a five-member high resolution ensemble, based on two global climate models (GCMs: ECHAM5 and CCCma3) and two regional climate models (RCMs: CLM and WRF) are analysed in this paper (Part II of a two part paper). In Part I the performance of the models for the control period are presented. The RCMs use a two nest procedure over Europe and Germany with a final spatial resolution of 7 km to downscale the GCM simulations for the present (1971–2000) and future A1B scenario (2021–2050) time periods. The ensemble was extended by earlier simulations with the RCM REMO (driven by ECHAM5, two realisations) at a slightly coarser resolution. The climate change signals are evaluated and tested for significance for mean values and the seasonal cycles of temperature and precipitation, as well as for the intensity distribution of precipitation and the numbers of dry days and dry periods. All GCMs project a significant warming over Europe on seasonal and annual scales and the projected warming of the GCMs is retained in both nests of the RCMs, however, with added small variations. The mean warming over Germany of all ensemble members for the fine nest is in the range of 0.8 and 1.3 K with an average of 1.1 K. For mean annual precipitation the climate change signal varies in the range of ?2 to 9 % over Germany within the ensemble. Changes in the number of wet days are projected in the range of ±4 % on the annual scale for the future time period. For the probability distribution of precipitation intensity, a decrease of lower intensities and an increase of moderate and higher intensities is projected by most ensemble members. For the mean values, the results indicate that the projected temperature change signal is caused mainly by the GCM and its initial condition (realisation), with little impact from the RCM. For precipitation, in addition, the RCM affects the climate change signal significantly.