Recharge on Non- irrigated Land Mike McVay 06/24/2010
Recharge on Non-irrigated Land
Mike McVay06/24/2010
Our Process for Estimating Recharge.
Our process for estimating recharge on non-irrigated land is to use ET Idaho data from 45 weather stations to calculate the amount of recharge based on precipitation, root zone moisture and soil type. Then we utilize semi-log kriging to interpolate the recharge data across the model domain.
Interpolation – Data Transformation
Both Willem and I are employing semi-log kriging to interpolate the data. This involves a semi-log transform of the data in which values below a threshold are transformed logarithmically and those values above the threshold are transformed linearly.
•Transformation of data is necessary to avoid negative extrapolated values.•Semi-log transformation employed.
•Prevents extrapolation of “small” numbers to negative values while maintaining linear behavior of “large” numbers. •Threshold (α) of 0.001.
Interpolation - Kriging
•I have employed Geostatistical Analyst in ArcMap to Krige, Willem has written code.
•Ordinary Kriging – No trend in data, so no drift has been incorporated for interpolation.•No nugget has been employed.•Linear Variogram Model – Few data points over a large area precludes fitting a variogram model to data.
•Geostatistical Analyst does not support Linear Variogram, so a Spherical model has been employed with a very large range.
•Sites with no data have been ignored during Kriging.
Interpolation Comparison
I have randomly chosen 30 month/soil-type combinations to compare interpolation results. My comparison of the results indicates that the Kriging methods that Willem and I have
employed are very similar (see table on next slide).
Stress Period/Soil Type McVay Sum (ft)
Schreuder Sum (ft)
Diff (%McVay
Sum)
Abs Diff (%McVay
Sum)Dec 2001 Lava 1455.88 1455.67 0.01% 0.0001Nov 1984 Thin 1149.73 1150.38 -0.06% 0.0006Dec 1998 Thin 459.68 459.99 -0.07% 0.0007May 1980 Thin 1639.82 1638.59 0.08% 0.0008Jan 1982 Lava 781.20 780.42 0.10% 0.0010Dec 2004 Lava 2158.85 2156.18 0.12% 0.0012May 1993 Lava 1271.83 1269.69 0.17% 0.0017Sep 1993 Thick 49.54 49.62 -0.17% 0.0017Sep 1997 Thin 183.48 183.80 -0.17% 0.0017Aug 1999 Thin 76.12 76.26 -0.18% 0.0018Sep 1994 Thick 27.85 27.79 0.21% 0.0021Aug 1989 Thick 38.84 38.76 0.22% 0.0022Jun 2003 Thin 22.65 22.74 -0.36% 0.0036Feb 1990 Lava 311.58 310.33 0.40% 0.0040Oct 2002 Thick 65.36 65.10 0.40% 0.0040Sep 1994 Lava 423.14 421.31 0.43% 0.0043Jul 1995 Thick 43.48 43.29 0.44% 0.0044Oct 2008 Lava 443.84 441.60 0.51% 0.0051Apr 1982 Thin 381.65 379.33 0.61% 0.0061May 2008 Lava 486.53 483.35 0.65% 0.0065Apr 1994 Lava 465.45 462.40 0.66% 0.0066Jan 1995 Thick 74.22 73.73 0.66% 0.0066Oct 1984 Thick 166.46 165.25 0.73% 0.0073Mar 1988 Thick 67.78 67.09 1.02% 0.0102Oct 1999 Lava 60.54 59.89 1.09% 0.0109Jun 1986 Lava 335.28 339.81 -1.35% 0.0135
Mar 1997 Thick 69.00 67.98 1.47% 0.0147Dec 1995 Thick 85.58 84.09 1.74% 0.0174Jan 1992 Thin 34.79 34.10 1.97% 0.0197
Absolute Value Average
0.55%
Some figures to assess agreement between Kriging Efforts
McVay381.65
Schreuder379.33
Apr82THIN0.004473 - 0.021840
0.021841 - 0.038428
0.038429 - 0.062001
0.062002 - 0.097196
0.097197 - 0.175000
McVay465.45
Schreuder462.40
Apr94LAVA0.004990 - 0.030746
0.030747 - 0.043898
0.043899 - 0.062672
0.062673 - 0.090261
0.090262 - 0.147525
McVay38.84
Schreuder38.76
Aug89THICK0.000992 - 0.014334
0.014335 - 0.045670
0.045671 - 0.082803
0.082804 - 0.124821
0.124822 - 0.184267
Aug99THIN0.000998 - 0.006010
0.006011 - 0.015079
0.015080 - 0.025996
0.025997 - 0.040262
0.040263 - 0.062212
McVay76.12
Schreuder76.26
Dec95THICK0.000996 - 0.003944
0.003945 - 0.009154
0.009155 - 0.016746
0.016747 - 0.035660
0.035661 - 0.071058
McVay85.58
Schreuder84.09
Dec98THIN0.001000 - 0.024633
0.024634 - 0.045599
0.045600 - 0.073201
0.073202 - 0.117129
0.117130 - 0.188134
McVay459.68
Schreuder459.99
Dec01LAVA0.001322 - 0.052426
0.052427 - 0.094935
0.094936 - 0.135564
0.135565 - 0.184064
0.184065 - 0.286403
McVay1455.88
Schreuder1455.67
Dec05LAVA0.000000 - 0.097000
0.097001 - 0.148000
0.148001 - 0.203000
0.203001 - 0.268000
0.268001 - 0.348063
McVay2158.85
Schreuder2156.18
Feb90LAVA0.001000 - 0.015000
0.015001 - 0.033000
0.033001 - 0.057000
0.057001 - 0.087000
0.087001 - 0.147086
McVay311.58
Schreuder310.33
Jan82LAVA0.000692 - 0.023813
0.023814 - 0.051942
0.051943 - 0.086757
0.086758 - 0.166045
0.166046 - 0.340169
McVay781.20
Schreuder780.42
Jan92THIN0.000994 - 0.004543
0.004544 - 0.010809
0.010810 - 0.030567
0.030568 - 0.066354
0.066355 - 0.125720
McVay34.79
Schreuder34.10
Jan95THICK0.000953 - 0.006299
0.006300 - 0.019629
0.019630 - 0.041309
0.041310 - 0.069145
0.069146 - 0.106479
McVay74.22
Schreuder73.73
Jul95THICK0.000997 - 0.006375
0.006376 - 0.018866
0.018867 - 0.034672
0.034673 - 0.052222
0.052223 - 0.081068
McVay43.48
Schreuder43.29
Jun86LAVA0.000985 - 0.019397
0.019398 - 0.038214
0.038215 - 0.060704
0.060705 - 0.086177
0.086178 - 0.112790
McVay335.28
Schreuder339.81
Jun03THIN0.000999 - 0.002957
0.002958 - 0.007500
0.007501 - 0.013213
0.013214 - 0.019579
0.019580 - 0.030036
McVay22.65
Schreuder22.74
Mar88THICK0.000998 - 0.003129
0.003130 - 0.006745
0.006746 - 0.011762
0.011763 - 0.018741
0.018742 - 0.034005
McVay67.78
Schreuder67.09
Mar97THICK0.000998 - 0.003515
0.003516 - 0.008089
0.008090 - 0.014215
0.014216 - 0.029570
0.029571 - 0.064268
McVay69.00
Schreuder67.98
May80THIN0.009725 - 0.076457
0.076458 - 0.128061
0.128062 - 0.170970
0.170971 - 0.221884
0.221885 - 0.341042
McVay1639.82
Schreuder1638.59
May93LAVA0.001128 - 0.056659
0.056660 - 0.111499
0.111500 - 0.167071
0.167072 - 0.224803
0.224804 - 0.312504
McVay1271.83
Schreuder1269.69
May08LAVA0.003000 - 0.024000
0.024001 - 0.040000
0.040001 - 0.062000
0.062001 - 0.090000
0.090001 - 0.135000
McVay486.53
Schreuder483.35
Nov84THIN0.000000 - 0.026000
0.026001 - 0.070000
0.070001 - 0.112000
0.112001 - 0.158000
0.158001 - 0.341382
McVay1149.73
Schreuder1150.38
Oct84THICK0.000996 - 0.006623
0.006624 - 0.014086
0.014087 - 0.024290
0.024291 - 0.038513
0.038514 - 0.065161
McVay166.46
Schreuder165.25
Oct99LAVA0.000999 - 0.003856
0.003857 - 0.006759
0.006760 - 0.010383
0.010384 - 0.017544
0.017545 - 0.031833
McVay60.54
Schreuder59.89
Oct02THICK0.000999 - 0.004149
0.004150 - 0.010469
0.010470 - 0.018581
0.018582 - 0.029108
0.029109 - 0.047011
McVay65.36
Schreuder65.10
Oct08LAVA0.000998 - 0.014497
0.014498 - 0.031277
0.031278 - 0.047166
0.047167 - 0.062931
0.062932 - 0.113931
McVay443.84
Schreuder441.60
Sep93THICK0.000994 - 0.006251
0.006252 - 0.029856
0.029857 - 0.072349
0.072350 - 0.125398
0.125399 - 0.190014
McVay49.54
Schreuder49.62
Sep94THICK0.000997 - 0.006222
0.006223 - 0.018034
0.018035 - 0.033891
0.033892 - 0.055521
0.055522 - 0.095243
McVay27.85
Schreuder27.79
Sep94LAVA0.000997 - 0.017574
0.017575 - 0.034539
0.034540 - 0.051538
0.051539 - 0.073862
0.073863 - 0.100641
McVay423.14
Schreuder421.31
Sep97THIN0.000000 - 0.006000
0.006001 - 0.017000
0.017001 - 0.030000
0.030001 - 0.053000
0.053001 - 0.095176
McVay183.48
Schreuder183.80
Calculating Recharge
Our Process for Calculating Recharge.
We calculate the amount of recharge at each of the 45 weather stations using ET Idaho precipitation (P), root-zone moisture (P_rz) data.
Recharge = P – P_rz
We represent recharge for the three general soil types of Thick Soil, Thin Soil and Lava Rock using the ET Idaho root-zone moisture data for Sage Brush and Dormant Turf covers.
To represent recharge on the various soil types we use:
1. Precipitation – Sage Brush Root-zone Moisture to represent Thick Soil.2. Precipitation – Dormant Turf Root-zone Moisture to represent Thin Soil.3. Precipitation – (2/3*Dormant Turf Root-zone Moisture) to represent Lava Rock.
What is P_rz?
“P_rz – Precipitation residing in the root zone. P_rz is the amount of gross reported precipitation that infiltrates into the soil (i.e. , less any runoff) and that remains in the root zone for consumption by evaporation or transpiration. P_rz is computed as:
P – Runoff – DPercwhere:
P is gross precipitationRunoff is estimated surface runoffDperc is deep percolation of any precipitation below the max root zone for the
land-use condition.”
Note: Definition from “Evaporation and Consumptive Irrigation Water Requirements for Idaho.” Allen and Robinson, 2007.
Note: All values are in units of [L/T] which are multiplied by the number of days in each month.
Our Proxy for Recharge.
Our proxy for recharge is calculated as:
Recharge = P – P_rz P-(P – Runoff – Dperc) Runoff + Dperc
Recharge = Runoff + Dperc
where:Recharge is recharge to the aquiferP is gross precipitationP_rz is precipitation residing in root zoneRunoff is estimated surface runoffDperc is deep percolation of any precipitation below the max root zone for the
land-use condition.
We assume that all runoff will eventually percolate via ponding or stream leakage to become recharge.
Root Zone
Precipitation
Runoff
P_rz
Dperc
Recharge = P – P_rz P-(P – Runoff – Dperc) Runoff + Dperc
Recharge = Runoff + Dperc
Three scenarios are present in the data from ET Idaho that confound our process for estimating recharge.
Recharge = P – P_rz Runoff + Dperc
There are three scenarios in the ET Idaho data that we need to address.1. Precipitation residing in root zone is negative.
i. Typically occurs during dry periods.ii. P – (-P_rz) results in abnormally large amount of recharge.
2. Precipitation residing in root zone is greater than gross precipitation.i. Typically occurs during non-summer months (one in July) – only 15 occurrences.ii. Difference is always 0.01 mm/day.iii. P – P_rz results in negative recharge value.
3. Dormant Turf precipitation residing in root zone is greater than that of Sage Brush.i. Occurs in 31% of data.ii. Does not appear to be seasonal.iii. P – P_rz results in more recharge on Thick Soils than on Thin Soils and Lava.
Scenario 1 – Root Zone Moisture is negative
Our work-around for moving forward is to consider this a soil moisture “deficit” and set recharge to be zero.
P_rz = P – Runoff – Dperc Negative values should not occur, based on mass balance.
American Falls August 1981: P – P_rzSage = 0.07 – (-2.06) = 2.13
Our work-around for moving forward is to compute recharge as the absolute difference .
P_rz = P – Runoff – Dperc P_rz > P should not occur, based on mass balance.
Scenario 2 – Root Zone Moisture is greater than precipitation
Swan Valley March 1985: P – PrzSage = 0.80 – 0.81 = -0.01
No work-around for moving forward has been employed.
Performing the P – P_rz calculation results in more recharge from Thick Soil than from Thin Soil and Lava.
Previous work has demonstrated that recharge from precipitation is partly a function of soil texture and depth, and the general relationships between the soil types represented on the ESPA are that recharge from Lava > Thin Soil > Thick Soil (Contor, 2004; Garabedian, 1992).
Scenario 3 – Dormant Turf P_rz is greater than Sage Brush P_rz
Blackfoot
ESPAM SoilsTHIN_CELLS
THICK_CELLS
LAVA_CELLS
May 1980 Recharge (ft)0.000 - 0.016
0.017 - 0.034
0.035 - 0.045
0.046 - 0.055
0.056 - 0.066
0.067 - 0.076
0.077 - 0.087
0.088 - 0.098
0.099 - 0.109
0.110 - 0.120
0.121 - 0.131
0.132 - 0.143
0.144 - 0.156
0.157 - 0.171
0.172 - 0.188
0.189 - 0.208
0.209 - 0.229
0.230 - 0.251
0.252 - 0.293
0.294 - 0.381
Sep 1997 Recharge (ft)0.000 - 0.002
0.003 - 0.004
0.005 - 0.007
0.008 - 0.010
0.011 - 0.013
0.014 - 0.016
0.017 - 0.020
0.021 - 0.023
0.024 - 0.027
0.028 - 0.030
0.031 - 0.035
0.036 - 0.040
0.041 - 0.046
0.047 - 0.053
0.054 - 0.062
0.063 - 0.073
0.074 - 0.088
0.089 - 0.104
0.105 - 0.123
0.124 - 0.143
ESPAM SoilsTHIN_CELLS
THICK_CELLS
LAVA_CELLS
Recharge Calculation Summary
•Both Willem and I are using the same methodology in calculating recharge. •There are three scenarios that confound our recharge proxy process.
•Precipitation residing in root zone is negative.•Precipitation residing in root zone is greater than gross precipitation.•Dormant Turf precipitation residing in root zone is greater than that of Sage Brush.
•We have employed two temporary work-around solutions to move forward with interpolation.
•Precipitation residing in root zone is negative. Work-around: Recharge = Zero.•Precipitation residing in root zone is greater than gross precipitation. Work-around: Recharge = 0.01 mm/day.•Dormant Turf precipitation residing in root zone is greater than that of Sage Brush. Work-around: None.
•Formal solutions should be pursued by the committee once Dr. Allen has had a chance to look at the data.
Any Questions?