Hamidreza Norouzi

Systematic and random error components in satellite precipitation data sets

GEOPHYSICAL RESEARCH LETTERS, VOL. 39, L09406, doi:10.1029/2012GL051592, 2012 Systematic and random error components in satellite precipitation data sets Amir AghaKouchak, 1 Ali Mehran, 1 Hamidreza Norouzi, 2 and Ali Behrangi 3 Received 7 March 2012; revised 13 April 2012; accepted 13 April 2012; published 11 May 2012. [ 1 ] This study contributes to characterization of satellite precipitation error which is fundamental to develop uncertainty models and bias reduction algorithms. Systematic and random error components of several satellite precipitation products are investigated over different seasons, thresholds and temporal accumulations. The analyses show that the spatial distribution of systematic error has similar patterns for all precipitation products. However, the systematic (random) error of daily accumulations is significantly less (more) than that of high resolution 3-hr data. One should note that the systematic biases of satellite precipitation are distinctively different in the summer and winter. The systematic (random) error is remarkably higher (lower) during the winter. Furthermore, the systematic error seems to be proportional to the rain rate magnitude. The findings of this study highlight that bias removal methods should take into account the spatiotemporal characteristics of error as well as the proportionality of error to the magnitude of rain rate. Citation: AghaKouchak, A., A. Mehran, H. Norouzi, and A. Behrangi (2012), Systematic and random error components in satellite precipitation data sets, Geophys. Res. Lett., 39, L09406, doi:10.1029/2012GL051592. 1. Introduction [ 2 ] Over the past three decades, development of satellite sensors have resulted in multiple sources of precipitation data sets. However, the quantification and understanding of uncertainties associated with remotely sensed satellite data remains a challenging research topic Bellerby and Sun [2005]. The uncertainties of satellite precipitation data arise from different factors including the sensor itself, retrieval error, and spatial and temporal sampling, among others [e.g., Hong et al., 2006]. [ 3 ] Numerous studies have addressed validation, verifi- cation and uncertainty of satellite precipitation estimates against ground-based measurements [e.g., Turk et al., 2008; Ebert et al., 2007]. This study aims to go beyond the vali- dation and inter-comparison of satellite products by analyz- ing error characteristics of precipitation algorithms. In this Department of Civil and Environmental Engineering, University of California, Irvine, Irvine, California, USA. Department of Construction Management and Civil Engineering Technology, City University of New York, New York City College of Technology, Brooklyn, New York, USA. Jet Propulsion Laboratory, California Institute of Technology, Pasadena, California, USA. paper, systematic and random error components of several satellite precipitation products are investigated over different seasons, thresholds and temporal accumulations. Ideally, the systematic error is to be removed or minimized. In mea- surement theory, many algorithms have been developed to reduce systematic error with the aim of reducing the overall uncertainty Taylor [1999]. Evidently, understanding error properties including systematic and random components are fundamental for future improvements in precipitation retrieval algorithms, development of uncertainty models and bias adjustment techniques, and many other research studies and operational applications [Sorooshian et al., 2011]. 2. Data Resources [ 4 ] The following satellite precipitation data sets are used for error analysis: (a) The CPC MORPHing (CMORPH) [Joyce et al., 2004] algorithm; (b) The Precipitation Esti- mation from Remotely Sensed Information using Artificial Neural Networks (PERSIANN) [Sorooshian et al., 2000]; (c) The Tropical Rainfall Measuring Mission (TRMM) Multi-satellite Precipitation Analysis (TMPA) real-time (hereafter, 3b42-RT) [Huffman et al., 2007]. [ 5 ] The Stage IV radar-based gauge-adjusted precipita- tion data, available from the National Center for Environ- mental Prediction (NCEP), are used as the reference data set. The Stage IV data include merged operational radar data and rain gauge measurements in hourly accumulations and 4 km grids. The Stage IV observations are accumulated to 3-hourly and aggregated onto 0.25 grids to match with satellite data. The study area covers the entire conterminous United States (hereafter, CONUS). Three years of precipi- tation data (01/01/2005–12/31/2007) are used for the anal- ysis. Hereafter, the difference between satellite estimates and Stage IV observations is termed as precipitation error. 3. Methodology and Results [ 6 ] In this study, the Willmott decomposition technique is used for deriving the systematic and random components of error. Willmott [1981] suggested that the error in the numerical weather prediction models can be separated into systematic and random error components as: n A X n A X n ¼ i¼1 i¼1 Corresponding author: A. AghaKouchak, Department of Civil and Environmental Engineering, University of California, Irvine, CA 92617, USA. (amir9a@gmail.com) Copyright 2012 by the American Geophysical Union. 0094-8276/12/2012GL051592 P sat A P ref P* sat A P ref n A X i¼1 P sat A P* sat n n where: P sat = satellite estimates P ref = reference measurements (here, Stage IV) L09406 1 of 4

Quantifying Uncertainties in Land-Surface Microwave Emissivity Retrievals

Uncertainties in the retrievals of microwave land-surface emissivities are quantified over two types of land surfaces: desert and tropical rainforest. Retrievals from satellite-based microwave imagers, including the Special Sensor Microwave Imager, the Tropical Rainfall Measuring Mission Microwave Imager, and the Advanced Microwave Scanning Radiometer for Earth Observing System, are studied. Our results show that there are considerable differences between the retrievals from different sensors and from different groups over these two land-surface types. In addition, the mean emissivity values show different spectral behavior across the frequencies. With the true emissivity assumed largely constant over both of the two sites throughout the study period, the differences are largely attributed to the systematic and random errors in the retrievals. Generally, these retrievals tend to agree better at lower frequencies than at higher ones, with systematic differences ranging 1%-4% (3-12 K) over desert and 1%-7% (3-20 K) over rainforest. The random errors within each retrieval dataset are in the range of 0.5%-2% (2-6 K). In particular, at 85.5/89.0 GHz, there are very large differences between the different retrieval datasets, and within each retrieval dataset itself. Further investigation reveals that these differences are most likely caused by rain/cloud contamination, which can lead to random errors up to 10-17 K under the most severe conditions.

Global Land Surface Emissivity Estimation From AMSR2 Observations

A reliable estimate of emissivity is critical for a wide range of applications for the atmosphere, the biosphere, the lithosphere, the cryosphere, and the hydrosphere. This study uses three years (August 2012 to July 2015) of data from the Advanced Microwave Scanning Radiometer-2 sensor that is onboard the Global Change Observation Mission 1st Water satellite to explore estimates of instantaneous global land emissivity. A method is adopted to remove the known inconsistency in penetration depths between microwave brightness temperatures and infrared-based ancillary data that could cause differences between day and night emissivity estimates. After removing the diurnal atmospheric effects, the resulting retrieved cloud-free land emissivities realistically represent well-known large-scale features. As expected, the polarization differences of estimated emissivities show noticeable seasonal variations over the deciduous woodland and grassland regions due to changes in vegetation density. The potential of estimated emissivities for high-latitude snow detection and freeze/thaw state identification is also demonstrated.

Urban surface energy budget study using flux tower observations and remote sensing measurements

Urban heat islands cause that built up areas experience warmer temperature than their surrounding rural regions. This issue can adversely affect the energy consumption and public health especially in highly populated cities. The aim of this research study is to characterize the effect and the response of each surface type in the cities to increase our understanding of climate, anthropogenic heat, and urban heat islands. Flux tower observations as well as satellite-based remote sensing measurements are two source of valuable information. Flux towers are rarely deployed in the cities or built up environment and mostly take measurements in natural surfaces. Here we deploy several flux towers on different surface in New York City to enhance our understanding about the reaction of each surface to the energy balance. Complete energy balance stations are installed over distinct materials such as concrete, asphalt, and rooftops. This study can help to provide a novel approach to use ground observations and map the maxima and minima air temperature in New York City using satellite measurements. Satellites also provide many measurements from the earth surface at various spatial resolutions. MODIS data sets particularly deliver skin temperature. Moreover, satellite observations from Landsat 8 are utilized to classify the city surfaces to distinct defined surfaces where ground observations were obtained. The mapped temperatures will be linked to MODIS surface temperatures to develop a model that can downscale MODIS skin temperatures to fine resolution air temperature over urban regions. The evaluation of results against independent ground observations reveals that the proposed method is promising for studying surface energy balance in urban regions.

Fine temporal resolution freeze and thaw states using combination of microwave land surface emissivity estimated

Monitoring freeze-thaw (FT) transitions in high latitude regions are critical to enhancing our knowledge about the prediction of biogeochemical transitions, carbon dynamics, climate change, and impacts on boreal-arctic ecosystems. Since land surface emissivity depends primarily on the surface characteristics, it would contains valuable information about the surface, especially regarding freeze and thaw states. The surface characteristics in terms of microwave emission changes whenever water undergoes phase changes at constant temperature. This study aims to investigate the potential of using emissivity estimates from various microwave sensors such as the Advanced Microwave Scanning Radiometer — Earth Observing System (AMSR-E), Special Sensor Microwave Imager (SSM/I), AMSR2, and the Global Precipitation Measurement (GPM) Microwave Imager (GMI). It employs data fusion techinques to construct diurnal estimates in order to accurately predicting the exact time of the freeze-thaw transition for each land cover type and region. The results reveal that emissivity difference values from low and high frequencies (such as 6.9GHz and 89GHz) at horizontal polarization have a strong correlation with ground-based soil temperature values at 5cm depth. A novel threshold-based approach specific to different land cover types is proposed for daily FT detection from the use of three years (August 2012–July 2015) of emissivity estimates at different frequencies. Ground-based soil temperature observations are used as reference to develop threshold values for FT states. Preliminary evaluation of the proposed approach with independent ground observations for the year 2015 shows that the use of land emissivity estimates for high-latitude FT detection is promising with fine temporal resolution (at least 4 times a day).

Bridging Nongeoscience STEM Majors to the Geoscience Workforce through a Skills Training and Enrichment Program

AbstractNew York City College of Technology has created a year-round geoscience workforce preparation and geoscience career mentoring program for nongeoscience, minority science, technology, engineering, and math (STEM) students beginning at the critical juncture of their junior year. The overall goal of the program is to create a viable pathway to the geoscience workforce by tapping into a nontraditional pool of students. Each year 12 students are recruited to participate in a structured geoscience workforce model program that consists of geoscience exposure, preparation, apprenticeship, and experience. The students not only receive support with cohort-building activities, but they also participate in two geoscience internship programs that equip them with geoscience knowledge; geoscience workforce skills; summer internships at a federal, local, or private geoscience facilities; mentoring by geoscience practitioners; and networking opportunities with geoscience companies and geoscience professional socie...

Equipping undergraduate STEM majors with Geoscience and remote sensing tools: A pathway to replenishing the Geoscience workforce

Reversing the dramatic decrease in geoscience interest, awareness, participation, and preparation among students at all levels in the United States has become a paramount priority, particularly since recent studies ([1], [2], and [3]) project that this critical regression is expected to continue well into this young century. Investments by the U.S. National Science Foundations Opportunities for Enhancing Diversity in the Geosciences program ([4], [5], [6], [7], [8], [9]) have yielded innovative insights, practical strategies, and replicable models that are designed to broaden student access, participation, and success at various stages of the geoscience pipeline. However, despite these transformative initiatives, more needs to be done to ensure that evidence-based practices are utilized to attract students and to increase their persistence and achievement in the geosciences. At the New York City College of Technology (City Tech) of the City University of New York, a new, pioneering, piloted pathway model in which undergraduates majoring in STEM disciplines are being equipped with vital geoscience and remote sensing skills for the geoscience workforce is successfully on-going. Components of this work are presented in this manuscript.

High-latitude freeze and thaw states detection using satellite-based microwave land surface emissivity estimates

Freeze and thaw (FT) processes have profound impact on the terrestrial water cycle, net primary productivity, carbon cycle, surface energy budget and hence the global climate system. The available passive microwave based FT states data are basically developed from brightness temperatures, which themselves affected by atmospheric water vapor content. Since land surface emissivity estimates derived from passive microwave observations are free from atmospheric effects, the use of land emissivity in FT states detection is promising. The objective of this study is to estimate land surface emissivity from the Advanced Microwave Scanning Radiometer-2 observations and to investigate its potential for high-latitude FT states detection. The instantaneous land surface emissivity is computed using an improved algorithm along with near-simultaneous ancillary data sets. The difference of estimated land emissivity between higher and lower frequency channels shows great potential for FT states detection.

Consistency analysis among microwave land surface emissivity products to improve GPROF precipitation estimations

To understand the atmospheric phenomena such as rain rate, cloud liquid water, and total precipitable water from satellite microwave observations, the surface contribution should be accounted and be removed from the microwave signal. The objective of this proposed research is to develop a land surface emissivity that facilitates providing this information. The emissivity product will improve the Goddard PROFiling algorithm (GPROF) precipitation estimates. It makes use of microwave measurements from newly launched Global Precipitation Mission (GPM) Microwave Imager (GMI) sensor to produce an emissivity database for a range of frequencies from 6.9 GHz (C band) to high frequencies such as 183 GHz. The goal of this work is to inter-compare four global land surface emissivity products over various land-cover conditions to assess their consistency. The intercompared retrieved land emissivity products were generated over five-year period (2003-2007) using observations from the Advanced Microwave Scanning Radiometer - Earth Observing System (AMSR-E), Special Sensor Microwave Imager (SSM/I), The Tropical Rainfall Measuring Mission (TRMM) Microwave Imager (TMI) and Windsat. First, all products were reprocessed in the same projection and spatial resolution as they were generated from sensors with various configurations. Then, the mean value and standard deviations of monthly emissivity values were calculated for each product to assess the spatial distribution of the consistencies/inconsistencies among the products across the globe. The emissivity products were also compared to soil moisture estimates and satellite-based vegetation index to assess their sensitivities to the changes in land surface conditions.

Evaluation of radar precipitation estimates near gap regions: a case study in the Colorado River basin

Radar precipitation estimation is very useful for hydrological and climatological studies. However, radar precipitation has inherent difficulty in estimating precipitation in mountainous regions. In developed countries such as the United States where there are extensive precipitation radar networks, gaps in the radar precipitation field are usually due to radar beam blockage by mountains. The goal of this study is to evaluate the performance of a daily radar precipitation field (Stage-II) against rain gauge measurements near radar gap areas in the Colorado River basin of the United States (southwestern Colorado, southeastern Utah, northeastern Arizona and northwestern New Mexico). We evaluated daily precipitation data for the years spanning from 2007 to 2009. Statistical score skills including correlation and bias are used for evaluation. Compared to gauge measurements, Stage-II fails to capture the altitude dependence of precipitation in the region. Bias analysis shows that Stage-II underestimates precipitation at higher elevation. Seasonal evaluations of Stage-II indicate that it underestimates cold season precipitation in the study area. Overall, the results show that the error in Stage-II precipitation estimates made within 100 km from the gap area, as measured against rain gauge measurements, is considerable, and caution is warranted for its use in hydrological and water management applications.