There are three hydrology data sets, all of which contain measurements of stream flow from the northwest quadrant of the FIFE site. This quadrant of the FIFE site contains the The Konza Prairie Long-Term Ecological Research (LTER) site. The data in the three data sets was collected by three organizations, the LTER staff, the USGS, and the investigator group of Wang, et al. Four of the six stations in these data sets were or are operated by the LTER site. These four stations are placed on four tributaries to Kings Creek, the primary stream in the LTER site. The USGS operated one station on Kings Creek about 2.9 miles from the mouth of the creek. The sixth station was built and operated by Ted Engman (then with USDA) as part of the FIFE investigation led by Jim Wang. Their station was positioned at the southern end of watershed 1D, which is part of the Swede Creek drainage. Both Kings Creek and Swede Creek are tributaries of McDowell Creek, which ultimately drains into the Kansas River. A visual representation of these watersheds and the location of the stations on each creek is in the Geographic Reference data set found on FIFE CD-ROM Volume 5.

The LTER stations had permanently installed V-flume weirs from which the water flow was monitored and measured. The station on watershed 1D had a temporary V-notched, sharp-crested, cross-channel weir installed for the duration of the experiment. The permanent USGS station did not have a weir installed; rather, it relied upon the geometry of the Kings Creek channel. At the location of the monitoring station, the streambed was a series of limestone terraces which were frequently cleared of debris by the flowing water. At all six stations the water flow was monitored indirectly by measuring the height of the water in the creek using a stilling well.

The three data sets in this group, daily stream flow , 15-minute stream flow, and stream flow during storms, all contain measurements of volumetric stream flow. The daily and 15-minute data sets differ in that the daily data is for five of the six stations, the four LTER stations and the USGS station, while the 15-minute data is only for the USGS station on Kings Creek. The storm flow data on the other hand is also for five stations but this data set has the data from the 1D watershed rather than from the USGS station.

Collectively these data sets are useful as an element of the equation used to compute water balance for this portion of the FIFE site. The other elements such as precipitation, soil moisture, evaportranspiration, etc., that are necessary to compute water balance are included in other data sets on these FIFE CD-ROMs. These data sets, especially those by the LTER site, are also useful in comparing the effects of vegetation and land treatments on the rate and extent of runoff. The four stations managed by the LTER site are in four different subwatersheds which receive different land treatments (grazing, burning, and fertilization) and as a consequence have different vegetation cover.

This data set contains measurements of stream flow volume from small watersheds during storm events. Measurements for all other times are in the Stream Flow Daily data set or are not available. Four of the weirs are permanently maintained by the Konza Prairie LTER program and the fifth was installed for the duration of the FIFE field experiment, also within the Konza Prairie research area. The LTER weirs were on tributaries of Kings Creek. These data cover the period from January 1, 1987 through January 1, 1988. During the storm events measurements were recorded at most every five minutes.

This data set contains the average daily stream flow for each of the permanent weirs installed by the LTER staff within the Konza Prairie LTER site, located within the northwest quadrant of the FIFE study area, and for the USGS station on Kings Creek. The four LTER stations are in different subwatersheds, but all of the tributaries feed into Kings Creek above the point of the USGS station. The minimum and maximum flow rates for each day are also included as well as the time that the minimum and maximum occurred. The data ranges from April 1, 1979 through September 2, 1988. However, not all stations report data for this entire period. The USGS station reports data for the entire period but most of the LTER stations report data for only June 1985 through December 1987. (Note that 1988 was a drought year, and there may have been little flow at the LTER stations).

These data were collected at the USGS station on Kings Creek and have a resolution of at most 15 minutes. The resolution can be as much as 24 hours if there is little to no water flow in the creek. The data range from April 1979 to September 1988. The data set contains the average flow for the specified period.

A major asset of these data is their close spatial and temporal relationship. They are all from the same quadrant of the FIFE site, thereby providing detailed runoff information for a reasonably large spatial area, approximately 7 km x 7 km. Moreover, they are supported by a large number of soil moisture and rainfall measurements for this same quadrant and for the same periods.

These data, because they were collected by three different groups have inherent differences due to different instrumentation techniques. In particular, two of the groups (LTER and Wang, et al.) used weirs positioned within the stream while the USGS station did not use a weir. Moreover, the groups that used weirs, used two very different types. The weirs used by the LTER group were embedded in the stream bottom, while that used by Wang, et al. was installed across the channel of the stream. These inherent differences may create discrepancies in these data since the accuracy and reliability of the water flow estimates vary with these techniques.

These data sets represent all of the data sets that contain measurements of runoff for the FIFE. These data are routine monitoring data, yet they are essential if water balances or effects of landscape upon runoff are to be understood. The 15-minute data set is especially useful for water balances because it has sufficient temporal resolution to capture short-term changes in the stream flow that are related to short-lived intermittent storm events. The daily data set, on the other hand, provides the spatial dimension, since this data is from five different gaging locations. Lastly, the water flow during storm events provides sufficient detail from five watersheds to give a fairly complete indication of the runoff related to storm events and the stream flow response to pulses of water from storm events. Both the Daily and Storm data sets are comprised primarily of data collected at the Konza Prairie LTER weirs, while the USGS data contributed to the Daily data set and is the sole station in the 15-minute data set. The overlapping stations in the different data sets helps to link them together and to provide links between the high frequency (15 minutes) and low frequency (daily) data sets, and between storm event measurements and normal flow.

These data sets are related to several other data sets which contain data for other components of the hydrologic cycle. Those FIFE data sets which contain rainfall, soil moisture, and soil properties are all related to the data sets in this group. These data sets are described in the Surface Meteorological Data, Soil Properties Data, and Soil Moisture Data Group Summary documents. Several data sets in the Surface Meteorological data group contain measurements of rainfall, humidity, and cloud coverage, while the Soil Properties Data Group has data sets that describe the water holding properties of the soil within the FIFE site, and the Soil Moisture Data Group has several data sets that measure the moisture contained with the soil at different times throughout the FIFE site. The data in these data sets are all related in that they are necessary components to compute a water balance for the FIFE site.

These data sets are generally of high quality. The procedures used to collect them are well known and well understood. This is especially true for the data collected at the LTER stations and the USGS station. The least confidence resides with the data collected on watershed 1D since the methods used to collect these data and the equations used to convert the raw measurements to stream flow are unknown. The USGS data, although thought to be of high quality, suffers from a lack of adequate information about the processing equations. However, USGS has standard procedures for computing stream flow (which differ between reporting stations) which are available from USGS with some effort. Finally, the data from the LTER stations is most prone to error during storm flow since these weirs were not calibrated at storm flow water levels.

Caution should be used in relying upon these data to try to complete a water balance. The FIFE study area is underlain by alternating layers of limestone and shale (see Geology information on CD-ROM Volume 5), creating ample potential for subsurface drainage and flow. During periods of low flow, the small tributaries in the Kings Creek watershed can be observed to disappear underground for portions of their runs. Such subsurface flows are not monitored.

These data were instrumental in producing a water balance model for the FIFE site that includes the major elements effecting the water balance of a catchment (Famiglietti, et al., 1992). This model incorporated evaporation, precipitation, runoff, changes in the storage of water in the unsaturated areas of the basin, and changes in ground water storage. Moreover, these data were instrumental in verifying that the model accurately described the migration of water through the components of the water balance equation. The results show the temporal and spatial distribution of the transition from an atmosphere-controlled water-saturated watershed to a soil-controlled unsaturated watershed.

Engman, E.T., G. Angus, and W. Kustas. 1989. Relationships between the hydrologic balance of a small watershed and remotely sensed soil moisture. Remote Sensing and Large Scale Processes, IAHS Publ. No. 186, Proc. IAHS, 3rd Int. Assoc., Baltimore, MD.

Wood, E.F. 1990. Water balance model for Kings Creek. Symposium on FIFE. Am. Meteorol. Soc., Boston, Massachusetts, pp. 163-166.

Engman, E.T., W. Kustas, T.J. Schmugge, and J.R. Wang. 1987. Relationship among the remotely sensed soil moisture, streamflow, and evapotranspiration. AGU Fall Meeting, San Francisco.

Sivapalan, M., K.J. Beven, and E.F. Wood. 1987. On hydrologic similarity: 2. A scaled model of storm runoff production. Water Resource Res. 23:2266-2278.

Wood, E.F., M. Sivapalan, K.J. Beven, and L. Band. 1989. Effects of spatial variability and scale with implications to hydrologic modeling. J. Hydrol. 102:29-47.

Famiglietti, J.S., E.F. Wood, M. Sivapalan, and D.J. Thongs. 1992. A catchment scale water balance for FIFE. J. Geophysical Res. 97(D17):18997-19007.