Skip to main content
ARS Home » Pacific West Area » Tucson, Arizona » SWRC » Research » Publications at this Location » Publication #312206
Title: Models for estimating daily rainfall erosivity in China

Author
item XIE, Y. - Beijing Normal University
item YIN, S.Q. - Beijing Normal University
item LIU, B. - Beijing Normal University
item Nearing, Mark
item ZHAO, Y. - University Of Arizona

Submitted to: Journal of Hydrology
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 2/14/2016
Publication Date: 2/20/2016
Citation: Xie, Y., Yin, S., Liu, B., Nearing, M.A., Zhao, Y. 2016. Models for estimating daily rainfall erosivity in China. Journal of Hydrology. 535:547-558. https://doi.org/10.1016/j.jhydrol.2016.02.020.
DOI: https://doi.org/10.1016/j.jhydrol.2016.02.020

Interpretive Summary: Rainfall erosivity is an index that quantifies the power of a rainstorm to cause soil erosion. It is important because the rainfall erosivity index, or R-factor, is widely used around the world in estimation of soil erosion rates for purposes including design of conservation practices and assessment and inventory of local, regional, or national soil erosion rates. The R-factor is calculated using detailed rainfall information collected on at least five minute increments. Most rainfall data collected around the world is reported only on a daily basis. The purpose of this study was to look at detailed measured rainfall for more than 10,000 storms collected in China and develop or test existing statistically-based equations for relating daily rainfall amounts to rainfall erosivity. We looked at the possibility of estimating erosivity on time steps of daily values, half-month values (as is often used for conservation planning purposes), yearly, and long-term annual averages. Three different model types were assessed, two of which used only daily rainfall total values and one that also used the maximum 60 minute rainfall amount for a given day. We found that all three models worked reasonably well for estimating year-by-year and average annual erosivities. Models I and II did not work very well for estimating the day-by-day erosivities, but the third model, which used the 60 minute rainfall information, much improved those daily estimations. In general the third model worked better than the other two. We recommend using Model III in cases where 60 minute rainfall data is available, and using Models I and II only when estimating yearly or average annual erosivities.

Technical Abstract: The multiplication of rainfall energy and maximum 30 minutes intensity (EI30) is the most widely used rainfall erosivity index for empirical soil loss prediction models, however its calculation requires high temporal resolution rainfall data which are often not readily available in China in most locations. The purpose of this study was to develop models suitable for eastern China to estimate erosivity from a daily rainfall index, which is more widely available. One-minute resolution rainfall data derived from siphon, self-recording rain gauges for 16 stations with the erosivity R-factor ranging from 781.9 to 8258.5 MJ mm hm-2 h-1 yr-1 over the eastern water-erosion region of China were analyzed. A total of 5942 erosive events from one-minute resolution rainfall data of ten stations were used to develop three models. Another 4949 erosive events from the other six stations were used to validate the models. Results showed that 40.2% of days with erosive rainfall had a a single, entire rainfall event and 43.2% of those days had a rainfall event that passed across more than one day. The threshold of daily rainfall between events classified as erosive and non-erosive was suggested to be 10 mm based on these data. Parameters for three models were calibrated and validated. Two of the models (I and II) were based on a power law function that required only daily rainfall totals. The coefficient a for Model I was different in the cool season (Oct.-Mar.) and warm season (Apr.-Sept.), whereas a for Model II was fitted with a sinusoidal curve of seasonal variation. Both Model I and Model II estimated the erosivity index for average annual, yearly, and half-month temporal scales reasonably well, with the mean relative error ranging from 10.8% to 45.0%. Model II predicted slightly better than Model I. However, the prediction efficiency for daily erosivity index was limited, with the mean relative errors being 132.0% (Model I) and 118.5% (Model II) and Nash-Sutcliffe model efficiency being 0.55 (Model I) and 0.57 (Model II), when only daily rainfall amount was used. Model III, which used the combination of daily rainfall amount and daily maximum 60-min rainfall, improved predictions significantly, and produced a Nash-Sutcliffe model efficiency for daily erosivity index prediction of 0.93.