Evaluation of gene expression programming approaches for estimating daily evaporation through spatial and temporal data scanning

Evaporation of water from free water surfaces or from land surfaces is one of the main components of the hydrological cycle, and its occurrence is governed by various meteorological and physical factors. There is a multitude of models developed for estimating daily evaporation values by using weather data. This paper evaluates a Gene Expression Programming (GEP) model for estimating evaporation through spatial and temporal data scanning techniques. It is by using ‘leave‐one‐out’ procedures, a complete scan of the possible train and test set configurations is carried out according to temporal and spatial criteria. Comparison of the GEP model with empirical‐physical models shows that daily evaporation values computed by the GEP model are more accurate. Further, local calibration of the GEP model may not be needed, if enough climatic data are available at other stations. The performance of the GEP model fluctuates throughout the period of study and between stations. A suitable assessment of the model should consider a complete temporal and/or spatial scan of the data set used. Copyright © 2012 John Wiley & Sons, Ltd.     

Estimates of spatial variation in evaporation using satellite‐derived surface temperature and a water balance model

Evaporation dominates the water balance in arid and semi‐arid areas. The estimation of evaporation by land‐cover type is important for proper management of scarce water resources. Here, we present a method to assess spatial and temporal patterns of actual evaporation by relating water balance evaporation estimates to satellite‐derived radiometric surface temperature. The method is applied to a heterogeneous landscape in the Krishna River basin in south India using 10‐day composites of NOAA advanced very high‐resolution radiometer satellite imagery. The surface temperature predicts the difference between reference evaporation and modelled actual evaporation well in the four catchments (r² = 0·85 to r² = 0·88). Spatial and temporal variations in evaporation are linked to vegetation type and irrigation. During the monsoon season (June–September), evaporation occurs quite uniformly over the case‐study area (1·7–2·1 mm day^?1), since precipitation is in excess of soil moisture holding capacity, but it is higher in irrigated areas (2·2–2·7 mm day^?1). In the post‐monsoon season (December–March) evaporation is highest in irrigated areas (2·4 mm day^?1). A seemingly reasonable estimate of temporal and spatial patterns of evaporation can be made without the use of more complex and data‐intensive methods; the method also constrains satellite estimates of evaporation by the annual water balance, thereby assuring accuracy at the seasonal and annual time‐scales. Copyright © 2007 John Wiley & Sons, Ltd.     

Observation-Based Estimates of Global Glacier Mass Change and Its Contribution to Sea-Level Change

Glaciers have strongly contributed to sea-level rise during the past century and will continue to be an important part of the sea-level budget during the twenty-first century. Here, we review the progress in estimating global glacier mass change from in situ measurements of mass and length changes, remote sensing methods, and mass balance modeling driven by climate observations. For the period before the onset of satellite observations, different strategies to overcome the uncertainty associated with monitoring only a small sample of the world’s glaciers have been developed. These methods now yield estimates generally reconcilable with each other within their respective uncertainty margins. Whereas this is also the case for the recent decades, the greatly increased number of estimates obtained from remote sensing reveals that gravimetry-based methods typically arrive at lower mass loss estimates than the other methods. We suggest that strategies for better interconnecting the different methods are needed to ensure progress and to increase the temporal and spatial detail of reliable glacier mass change estimates.     

Estimation of potential evapotranspiration in the mountainous Panama Canal watershed

Spatially distributed hydrometeorological and plant information within the mountainous tropical Panama Canal watershed is used to estimate parameters of the Penman–Monteith evapotranspiration formulation. Hydrometeorological data from a few surface climate stations located at low elevations in the watershed are complemented by (a) typical wet‐ and dry‐season fields of temperature, wind, water vapour and pressure produced by a mesoscale atmospheric model with a 3 × 3 km² spatial and hourly temporal resolution, and (b) leaf area index fields estimated over the watershed during a few years using satellite data with two different spatial and temporal resolutions. The mesoscale model estimates of spatially distributed surface hydrometeorological variables provide the basis for the extrapolation of the surface climate station data to produce input for the Penman–Monteith equation. The satellite information and existing digital spatial databases of land use and land cover form the basis for the estimation of Penman–Monteith spatially distributed parameter values. Spatially distributed 3 × 3 km² potential evapotranspiration estimates are obtained for the 3300 km² Panama Canal watershed. Estimates for Gatun Lake within the watershed are found to reproduce well the monthly and annual lake evaporation obtained from submerged pans. Sensitivity analysis results of potential evapotranspiration estimates with respect to cloud cover, dew formation, leaf area index distribution and mesoscale model estimates of surface climate are presented and discussed. The main conclusion is that even the limited spatially distributed hydrometeorological and plant information used in this study contributes significantly toward explaining the substantial spatial variability of potential evapotranspiration in the watershed. These results also allow the determination of key locations within the watershed where additional surface stations may be profitably placed. Copyright © 2006 John Wiley & Sons, Ltd.     

Coupling SEBAL with a new radiation module and MODIS products for better estimation of evapotranspiration

ABSTRACT

Evapotranspiration (ET) is an important ecohydrological process especially in arid and semi-arid regions. In this study, a new radiation module based on MODIS data has been coupled with the Surface Energy Balance Algorithms for Land (SEBAL) to better estimate ET. The accuracies of the coupled model for estimating available energy and sensible heat (H) were improved significantly compared with the outputs from the original SEBAL which was based on empirical equations. The coupled SEBAL modelled instantaneous λET agreed much better with observations in the arid land of Central Asia than the original SEBAL, with a bias of ?2.86 W m^-2, root mean square error (RMSE) of 9.75 W m^-2, and normalized RMSE (NRMSE) of 0.13. The accuracy was blurred when scaling ET to a daily or monthly scale, mainly due to the uncertainties associated with temporal upscaling methods that were applied. Sensitivity analysis, which was conducted using numerical variance-based techniques, indicated that the estimated ET is sensitive to the available energy, suggesting the importance of obtaining accurate estimates of net radiation when applying the coupled SEBAL to estimate ET. This study provides a simple and reliable way to utilize MODIS products and contains sensitivity analysis for helping to correctly interpret the outputs, which are both important for large-scale ET estimation.     

A survey of studies into errors in large scale space-time averages of rainfall, cloud cover, sea surface processes and the earth's radiation budget as derived from low earth orbit satellite instruments because of their incomplete temporal and spatial coverage

This survey considers those studies conducted into estimating errors in satellite derived large scale space-time means (of the order of 250 km by 250 km by a month) for rainfall, cloud cover, sea surface processes and the Earth's radiation budget, resulting from their incomplete coverage of the space-time volume over which the mean is evaluated. Many of these studies have focused on estimating the errors in space-time means post satellite launch and compare mean data derived from such satellites with that from an independent data set. Pre-launch studies tend to involve computer simulations of a satellite overflying and sampling from an existing data set and hence the two approaches give values for sampling errors for specific cases. However, more generic sampling papers exist that allow the exact evaluation of sampling errors for any instrument or combination of instruments if their sampling characteristics and the auto-correlation of the parameter field are known. These generic and simulation techniques have been used together on the same data sets and are found to give very similar values for the sampling error and are presented. Also considered are studies in which data from several satellites, or satellite and ground based measurements are combined to improve estimates in the above means. This improvement being brought about not only by increased spatial and temporal coverage but also by a reduction in retrieval error.     

Continental Scale Hydrostratigraphy: Comparing Geologically Informed Data Products to Analytical Solutions

This study synthesizes two different methods for estimating hydraulic conductivity (K) at large scales. We derive analytical approaches that estimate K and apply them to the contiguous United States. We then compare these analytical approaches to three-dimensional, national gridded K data products and three transmissivity (T) data products developed from publicly available sources. We evaluate these data products using multiple approaches: comparing their statistics qualitatively and quantitatively and with hydrologic model simulations. Some of these datasets were used as inputs for an integrated hydrologic model of the Upper Colorado River Basin and the comparison of the results with observations was used to further evaluate the K data products. Simulated average daily streamflow was compared to daily flow data from 10 USGS stream gages in the domain, and annually averaged simulated groundwater depths are compared to observations from nearly 2000 monitoring wells. We find streamflow predictions from analytically informed simulations to be similar in relative bias and Spearman's rho to the geologically informed simulations. R-squared values for groundwater depth predictions are close between the best performing analytically and geologically informed simulations at 0.68 and 0.70 respectively, with RMSE values under 10 m. We also show that the analytical approach derived by this study produces estimates of K that are similar in spatial distribution, standard deviation, mean value, and modeling performance to geologically-informed estimates. The results of this work are used to inform a follow-on study that tests additional data-driven approaches in multiple basins within the contiguous United States.     

Intercomparison of the JULES and CABLE land surface models through assimilation of remotely sensed soil moisture in southeast Australia

Numerous land surface models exist for predicting water and energy fluxes in the terrestrial environment. These land surface models have different conceptualizations (i.e., process or physics based), together with structural differences in representing spatial variability, alternate empirical methods, mathematical formulations and computational approach. These inherent differences in modeling approach, and associated variations in outputs make it difficult to compare and contrast land surface models in a straight-forward manner. While model intercomparison studies have been undertaken in the past, leading to significant progress on the improvement of land surface models, additional framework towards identification of model weakness is needed. Given that land surface models are increasingly being integrated with satellite based estimates to improve their prediction skill, it is practical to undertake model intercomparison on the basis of soil moisture data assimilation. Consequently, this study compares two land surface models: the Joint UK Land Environment Simulator (JULES) and the Community Atmosphere Biosphere Land Exchange (CABLE) for soil moisture estimation and associated assessment of model uncertainty. A retrieved soil moisture data set from the Soil Moisture and Ocean Salinity (SMOS) mission was assimilated into both models, with their updated estimates validated against in-situ soil moisture in the Yanco area, Australia. The findings show that the updated estimates from both models generally provided a more accurate estimate of soil moisture than the open loop estimate based on calibration alone. Moreover, the JULES output was found to provide a slightly better estimate of soil moisture than the CABLE output at both near-surface and deeper soil layers. An assessment of the updated membership in decision space also showed that the JULES model had a relatively stable, less sensitive, and more highly convergent internal dynamics than the CABLE model.     

A review of the estimation of downward surface shortwave radiation based on satellite data: Methods,progress and problems

The estimation of downward surface shortwave radiation(DSSR) is important for the Earth's energy budget and climate change studies. This review was organised from the perspectives of satellite sensors, algorithms and future trends,retrospects and summaries of the satellite-based retrieval methods of DSSR that have been developed over the past 10 years. The shortwave radiation reaching the Earth's surface is affected by both atmospheric and land surface parameters. In recent years,studies have given detailed considerations to the factors which affect DSSR. It is important to improve the retrieval accuracy of cloud microphysical parameters and aerosols and to reduce the uncertainties caused by complex topographies and high-albedo surfaces(such as snow-covered areas) on DSSR estimation. This review classified DSSR retrieval methods into four categories:empirical, parameterisation, look-up table and machine-learning methods, and evaluated their advantages, disadvantages and accuracy. Further efforts are needed to improve the calculation accuracy of atmospheric parameters such as cloud, haze, water vapor and other land surface parameters such as albedo of complex terrain and bright surface, organically combine machine learning and other methods, use the new-generation geostationary satellite and polar orbit satellite data to produce highresolution DSSR products, and promote the application of radiation products in hydrological and climate models.     

A cascade approach to continuous rainfall data generation at point locations

High-resolution temporal rainfall data sequences serve as inputs for a range of applications in planning, design and management of small (especially urban) water resources systems, including continuous flow simulation and evaluation of alternate policies for environmental impact assessment. However, such data are often not available, since their measurements are costly and time-consuming. One alternative to obtain high-resolution data is to try to derive them from available low-resolution information through a disaggregation procedure. This study evaluates a random cascade approach for generation of high-resolution rainfall data at a point location. The approach is based on the concept of scaling in rainfall, or, relating the properties associated with the rainfall process at one temporal scale to a finer-resolution scale. The procedure involves two steps: (1) identification of the presence of scaling behavior in the rainfall process; and (2) generation of synthetic data possessing same/similar scaling properties of the observed rainfall data. The scaling identification is made using a statistical moment scaling function, and the log–Poisson distribution is assumed to generate the synthetic rainfall data. The effectiveness of the approach is tested on the rainfall data observed at the Sydney Observatory Hill, Sydney, Australia. Rainfall data corresponding to four different successively doubled resolutions (daily, 12, 6, and 3 h) are studied, and disaggregation of data is attempted only between these successively doubled resolutions. The results indicate the presence of multi-scaling behavior in the rainfall data. The synthetic data generated using the log–Poisson distribution are found to exhibit scaling behaviors that match very well with that for the observed data. However, the results also indicate that fitting the scaling function alone does not necessarily mean reproducing the broader attributes that characterize the data. This observation clearly points out the extreme caution needed in the application of the existing methods for identification of scaling in rainfall, especially since such methods are also prevalent in studies of the emerging satellite observations and thus in the broader spectrum of hydrologic modeling.     

A probability-weighted moment test to assess simple scaling

We present a statistically robust approach based on probability weighted moments to assess the presence of simple scaling in geophysical processes. The proposed approach is different from current approaches which rely on estimation of high order moments. High order moments of simple scaling processes (distributions) may not have theoretically defined values and consequently, their empirical estimates are highly variable and do not converge with increasing sample size. They are, therefore, not an appropriate tool for inference. On the other hand we show that the probability weighted moments of such processes (distributions) do exist and, hence, their empirical estimates are more robust. These moments, therefore, provide an appropriate tool for inferring the presence of scaling. We illustrate this using simulated Levystable processes and then draw inference on the nature of scaling in fluctuations of a spatial rainfall process.     

A probability-weighted moment test to assess simple scaling

We present a statistically robust approach based on probability weighted moments to assess the presence of simple scaling in geophysical processes. The proposed approach is different from current approaches which rely on estimation of high order moments. High order moments of simple scaling processes (distributions) may not have theoretically defined values and consequently, their empirical estimates are highly variable and do not converge with increasing sample size. They are, therefore, not an appropriate tool for inference. On the other hand we show that the probability weighted moments of such processes (distributions) do exist and, hence, their empirical estimates are more robust. These moments, therefore, provide an appropriate tool for inferring the presence of scaling. We illustrate this using simulated Levystable processes and then draw inference on the nature of scaling in fluctuations of a spatial rainfall process.     

Surface deformation induced by water influx in the abandoned coal mines in Limburg,The Netherlands observed by satellite radar interferometry

The coal reserves of Limburg, The Netherlands, have been exploited until the mid-1970's, leading to significant land subsidence, a large part of which was due to ground water pumping associated with the mining activities. In 1994, when also the hydrologically-connected neighboring German mining activities ceased, all pumps were finally dismantled. This resulted in rising groundwater levels in the mining areas, continuing until today. Here we report the detection and analysis of heterogeneous surface displacements in the area using satellite radar interferometry. The lack of adequate terrestrial geodetic measurements emphasizes the value of such satellite observations, especially in terms of the temporal and spatial characterization of the signal. Since the lack of direct mine water level measurements hampers predictions on future consequences at the surface, we study the relationship between surface deformation and sub-surface water levels in an attempt to provide rough correlation estimates and map the mine water dynamics.     

Regression method for estimating rainfall recharge at unconfined sandy aquifers with an equatorial climate

A reliable estimate of rainfall recharge is essential for groundwater system managements. This study develops a method based on regression equations for estimating rainfall recharge at unconfined sandy aquifers with an equatorial climate. The developed method (GR-I method) is generally efficient for estimating long-term regional recharge, as the computational procedures could be formulated and executed easily using Microsoft's Excel spreadsheet. More importantly, its application could be extended to sand textures different from the sand texture used in developing the regression equations. To evaluate its reliability, the method was applied to estimate monthly gross recharge percentages at the Changi reclaimed land. When ignoring the effect of rainfall clusters, the GR-I method was found to underestimate the monthly gross recharge percentages for those months with high monthly rainfall depths. By integrating the effect of rainfall clusters, the GR-I method yields reliable estimates of monthly gross recharge percentages. By including daily potential evaporation as an additional input variable, the Extended GR-I method was found to be not superior to the GR-I method, implying that soil moisture availability is the major governing factor for actual soil evaporation in the highly porous sand medium, instead of atmospheric demand represented by the potential evaporation rate. Using the GR-I method, the mean annual net recharge percentage of the study site was found to fall between 56·9 and 69·9%, which corresponds to a net recharge depth of 1073·8–1745·8 mm. Although the developed method provides a good alternative to other widely used methods, its recharge estimates still needs to be collaborated with estimates from other methods, as multiple techniques are highly recommended in any groundwater recharge estimations. Copyright © 2007 John Wiley & Sons, Ltd.     

Multifractal objective analysis: conditioning and interpolation

 We investigate various ways of statistically estimating multifractal fields from sparse data. First, the problem is set in the general framework of conditional expectations, and the effect of (multi) fractal sampling on the statistics of the measured process is investigated, showing how analytical expressions describing the statistical properties of the phenomenon should be modified by the sampling. Then, several techniques are introduced, our goal being to estimate the intensity of a field at resolution λ, given samples of the process collected by networks at higher resolutions Λ>λ. The general strategy underlying all the estimating techniques presented is to approximate the unknown field values at resolution λ by means of most likely estimates conditional to the available information at resolution Λ>λ. Finally, the procedures are tested on simulated lognormal multifractal fields sampled by means of fractal networks, and the propagation of the errors in a scaling framework is also discussed. These techniques are necessary for estimating geophysical processes in regions where no monitoring stations are present, a scenario often encountered in practice, and may also be of great help in studying natural hazards and risk assessment.     

Examination of Morton's CRAE model for estimating daily evaporation from field-sized areas

Morton's complementary relationship areal evapotranspiration (CRAE) model was originally designed to provide regional estimates of monthly evapotranspiration. Often, however, hydrologists and others require estimates of evapotranspiration for field-sized land units under a specific land use, for shorter intervals of time. This paper examines CRAE with respect to the algorithms used to describe different terms and its applicability to reduced spatial and temporal scales.

Daily estimates by CRAE of atmospheric radiation fluxes during the summer months are compared with monitored values. It is shown that errors in estimation of the extra-terrestrial flux, the transmittancy of clouds to short-wave radiation, the surface albedo and the net long-wave flux result in standard deviations of the difference between ‘modelled’ and ‘measured’ net all-wave radiation for 1-, 5- and 10-day periods of 2.58, 1.8 and 1.50 MJm⁻² day⁻¹ respectively.

The assumption in CRAE that the vapour transfer coefficient is independent of wind speed may lead to appreciable error in computing evapotranspiration. A procedure for incorporating a wind correction factor is described and the improvement in estimating regional evaporation is illustrated.

Comparisons of evapotranspiration estimates by CRAE and measurements obtained from soil moisture and precipitation observations in the semi-arid, cold-climate Prairie region of western Canada demonstrate that the assumptions that the soil heat flux and storage terms are negligible, lead to large overestimation by the model during periods of soil thaw.     

海冰反照率参数化和遥感反演方法及其产品的研究评述

反照率是控制地表能量收支的关键地球物理参数之一,海冰作为南北极地区重要的组成部分,海冰反照率的时空变化会对极地地区和全球范围的气候变化、物质平衡以及能量平衡等产生重要的影响.本文系统的总结了海冰反照率的影响因素、海冰反照率的参数化方法和遥感反演方法及产品的研究进展,阐明了各方法的基本原理、特点以及存在的问题等.海冰反照率的影响因素众多,主要受地表冰雪反射特性、太阳天顶角以及大气属性的影响;参数化方法提供了一种模拟海冰反照率的途径,主要通过取经验定值或建立温度、冰雪厚度以及光谱反照率等参数与海冰反照率的经验关系来进行,但是这种基于特定位置、特定时间以及特定的大气状态下的观测数据运用统计或经验方法建立的参数化方法,适用范围通常有限.遥感反演方法是高时空分辨率获取大范围及长时序海冰反照率的有力手段,主要分为传统的反演方法、直接反演方法以及基于非光学传感器的反演方法;但是遥感反演很容易受到云层的影响,仅能反演晴空下的海冰反照率,而且现有的方法基本都是针对单一传感器设计的,还没有能够联合多源传感器数据反演海冰反照率的方法.基于此,本文展望了未来海冰反照率的研究重点,即开展能够适用于云天空下的、高...     

Monitoring Continental Surface Waters by Satellite Altimetry

The monitoring of continental water stages is a requirement for meeting human needs and assessing ongoing climatic changes. However, regular gauging networks fail to provide the information needed for spatial coverage and timely delivery. Although the space missions discussed here were not primarily dedicated to hydrology, 18 years of satellite altimetry have furnished complementary data that can be used to create hydrological products, such as time series of stages, estimated discharges of rivers or volume change of lakes, river altitude profiles or leveling of in situ stations. Raw data still suffer uncertainties of one to several decimeters. These require specific reprocessing such as waveform retracking or geophysical correction editing; much work still remains to be done. Besides, measuring the flow velocity appears feasible owing to SAR interferometer techniques. Inundated surfaces, and the time variations of their extent, are currently almost routinely computed using satellite imagery. Thus, the compilation of the continuous efforts of the scientific community in these various investigative directions, such as recording from space the discharges of rivers or the change in water volume stored in lakes, can be foreseen in the near future.     

Comparative performance analysis of climate re-analysis approaches in Angola

This paper examines the performance of three climate re-analysis datasets in Angola, namely the ERA-40, NCEP-r1 and JRA-55, by comparing annual and seasonal estimates of precipitation, surface air temperature and evaporation with ground observation measurements. The observational dataset describes a region poorly covered by international climate databases and it is believed that most of its data have not been used in the data assimilation procedures of the climate models. This paper therefore provides a fresh look at the performance of these climate re-analysis datasets in a vast area where distance and civil war have hindered ground monitoring efforts. The re-analysis exercises offer better temperature estimates than precipitation estimates. When offered, the evaporation estimates from all three products are very poor. The three products are able to describe the main features of the spatial distribution of average annual precipitation and temperature, but struggle to reproduce the temporal changes of these variables. The results from a set of performance criteria show that the correlation between the observed ground measurements and re-analysis estimates is poor overall and the NSE values indicate that the average measured value at each location is usually a better estimate than the re-analysis estimate.

EDITOR A. Castellarin

ASSOCIATE EDITOR S. Kanae     

The influence of Gulf of Mexico Loop Current intrusion on the transport of the Florida Current

Based on an empirical orthogonal function analysis of satellite altimeter data, guidance from numerical model results, and CANEK transport estimates, we propose an index, based on differences in satellite-measured sea surface height anomalies, for measuring the influence of Gulf of Mexico Loop Current intrusion on vertically integrated transport variability through the Yucatan Channel. We show that the new index is significantly correlated at low frequencies (cut-off 120 days) with the cable estimates of transport between Florida and the Bahamas. We argue that the physical basis for the correlation is the geometric connectivity between the Yucatan Channel and the Straits of Florida.